beni mellal city morocco ... · cgiv2016 _ doctoral research workshop cgiv2016 forum - ii 8th...

i | CGiV2016

Barcelona ● Spain

© "Dark Days - Venice II" - Gabriele Peters

iV2015 & cgiv2016 - DIGITAL ART GALLERY Online Exhibition July 2015- June 2016

V I R T U A L G A L L E R Y V E N U E

www.g raph ic s l i nk .c o .uk /DART.h tm

Exhibiting Artists:

LiQin Tan

Anna Chupa

Joohyun Pyune

Santiago Echeverry

Harvey Goldman

Daniela Sirbu

John Corbett

Gabriele Peters

Nancy Wood

Heather Freeman

Bogdan Soba

Galt Tomasino

Gina Gibson

Margie Labadie

Dena Elisabeth Eber

Mohammad Majid al-Rifaie

Corinne Whitaker

Gloria DeFilipps Brush

Chris Kitchener

Matjuska Teja Krasek

Olivia Koval

Mike Kingan

Jean Constant

John Antoine Labadie

Alan Singer

Anna Ursyn

Neil Howe

Andras Szalai

Chiara Passa

Richard Merritt

© Joohyun Jung Pyune ~ Before Me

CGiV2016 _ Doctoral Research Workshop

CGiV2016 Forum - ii

8 t h D o c t o r a l R e s e a r c h W o r k s h o p Organised by

Computer Graphics, Imaging and Visualization Conference

In cooperation with

&

The Computer Graphics, Imaging and Visualization Conference (CGiV) is an international conference that aims to provide a foundation for integrating the Computer Graphics, Imaging and Visualisation in order to promote international exchange, cooperation and development. Building upon the reported success of last year’s workshop, CGiV is pleased to announce the

“8th Doctoral Research Workshop” which will run as part of the 13th International Conference on Computer Graphics, Imaging and Visualization (CGiV2016).

Doctoral Research workshop This workshop focuses on the issues that doctoral students face during their studies and includes following interactive

sessions – the theme for this year Research Impact for (PhD) researcher.

The workshop will be hands-on with short inputs from the instructor and time to work on the exercises and own projects where the instructor and researchers will operate as sparring partner in a friendly environment.

CGiV2016_PROGRAMME

1 | CGiV2016

Tuesday 29 March 2016 09:00 < Sultan Moulay Sliman University ● Faculty of Sciences and Technics ● Conference room hall >

Registration

10:00

-

13:0

< Sultan Moulay Sliman University ● Faculty of Sciences and Technics ● conference room >

Doctoral Research Workshop

Chairs: Banissi, Prof. Ebad, London South Bank University, UK

09:30 Registration

10.00 Opening & Welcome from discussion Panel members

10:10 Research Proposal

11:00 Student Project Discussion: “Using Grammars in Medical Imaging: A Comprehensive Overview“, Hamdi, Salah; Ben Abdallah,

Asma; Bedoui, Mohamed Hedi

A Spatial resolution and vacuum channeling phenomenon of the Neutron transmission through protection screen

Abderrazak Khanouchi

13:00 Research Impact

13:00 < Sultan Moulay Sliman University ● Faculty of Sciences and Technics ● Room101>

Lunch Break

14:00

-

17:00

< Sultan Moulay Sliman University ● Faculty of Sciences and Technics ● conference room>

Doctoral Research Workshop

14.00 Research Methodology

15:00 Student Project Discussion: Analysis and prediction of folding protein via the sequence of amino acids, Karbou, Saadia; El

Bahloul, Yasmina; Benaji, Brahim; Bouksaim, Mohamed; Taoudi Benchekroun, Majda

15:30 Break

16:00 Student Project Discussion:

Immunohistochemical analysis of some oncogenes of triple negative breast cancer (TNBC), Leouifoudi, Inass; SELLAMI, Souad;

HAKMAOUI, Abdelmalek; Zyad, Abdelmajid; RAIS, Hanae

Medical images segmentation using mean shift over Riemannian manifolds, GOUSKIR, Mohamed; BOUTALLINE, Mohammed;

BOUIKHALENE, Belaid; ELHADADI, Benachir

16:45 Group Discussion

17:00 Close

CGiV2016_PROGRAMME

2 | CGiV2016

Wednesday 30 March 2016

09:30 < Sultan Moulay Sliman University ● Faculty of Sciences and Technics ● conference room hall>

Registration 10:30 < Sultan Moulay Sliman University ● Faculty of Sciences and Technics ● conference room>

Opening & Welcome Conference Chair:

Prof. Bouchaib Mernarni, president of Sultan Moulay Sliman University, Morocco

Prof. Ahmed Zeghal, Dean of Faculty of Sciences and Technics, Sultan Moulay Sliman University, Morocco

Prof. Mohamed Fakir, Faculty of Sciences and Technics, Sultan Moulay Sliman University, Morocco

Prof. Muhammad Sarfraz, Department of Information Science, Kuwait University, Kuwait

Prof. Ebad Banissi, London South Bank University, UK

Prof. Pun, Chi Man, University of Macau, Macau S.A.R., China

Session CGiV2016_1.1: Visualisation Chair: Prof. Cherki Daoui, Sultan Moulay Sliman University, Morocco

Cloud Motion Estimation in Satellite Image Sequences by Tracking Skeleton Critical Points using Lucas-Kanade method

Hassan Id Ben Idder, Nabil Laachfoubi

Fire Event Analytics: A Visual Analytics Tool to Assist Wildland Fire Management

Nikolaos Georgoulopoulos, Ioannis Mariolis, Savvas Rogotis, Dimos Ioannidis, Konstantinos Votis, Dimitrios Tzovaras

11:15 < Sultan Moulay Sliman University ● Faculty of Sciences and Technics ● conference room hall >

Morning Coffee Break / Photograph Session 12:00 < Sultan Moulay Sliman University ● Faculty of Sciences and Technics ● conference room>

Session CGiV2014_1.2: Imaging Application Chair: Prof. Ahmad M. Aljamali, Kuwait University, Kuwait

Facial Feature Extraction and Recognition for Traditional Chinese Physiognomy Maolin Huang, Jessica Liu

Robust Region Descriptors for Shape Classification

Cong Lin, Chi Man PUN

Satellite Image Restoration by Applying the Genetic Approach to the Wiener Deconvolution Fouad Aouinti, M'barek Nasri, Mimoun Moussaoui, Soukaina Benchaou, Khalid Zinedine

CGiV2016_PROGRAMME

3 | CGiV2016


Lunch Break 14:20

- 15:20

< Sultan Moulay Sliman University ● Faculty of Sciences and Technics ●

conference room>

Session cgiv2016_1.3: Computer Graphics Chair: Dr Dimitrios Tzovaras, Center for Research and Technology

Hellas Information Technologies Institute, Greece

Local and Fast Fog Detection based on Saturation and RGB-Correlation around the Picture Vanshing Point Salma ALAMI, Abdelhak EZZINE, Fouad ELHASSOUNI

A stereo planar self-calibration method of a camera with variable intrinsic parameters

Ismail EL batteoui, Abderrahim Saaidi, Khalid Satori

Surface Tension and Wettability Modeling for Flowing Liquids

Mariusz Zubrzycki, Jacek Raczkowski

Attacks Against AODV Routing Protocol in Mobile Ad-Hoc Networks

Houda MOUDNI, Mohamed ER-ROUIDI, Hicham MOUNCIF, Benachir EL HADADI

< Sultan Moulay Sliman University ● Faculty of Sciences and Technics ●

Room102>

Session cgiv2016_1.4: Imaging Chair: Dr. Assma Azeroual, Computer Systems and Vision

Laboratory, Morocco

Phase extraction method in fringe projection profilometry using Instantaneous frequency

Rachid IDRISSI EL-AZAMI, Benaissa BELLACH, Hamid TAIRI

An Improvable Structure for Similarity Searching in Metric Spaces: Application on Image Databases

Youssef Hanyf, Hassan Silkan, Halima Labani

Face Recognition using Local Binary Probabilistic Pattern (LBPP) and 2D-DCT Frequency Decomposition

Dahmouni Abdellatif, Aharrane nabile, El Moutaouakil karim, Satori kalid

15:20 < Sultan Moulay Sliman University ● Faculty of Sciences and Technics ● conference room hall> Coffee Break

CGiV2016_PROGRAMME

4 | CGiV2016

15:50

- 17:00


Session cgiv2016_1.5: Imaging

Chair: Prof. Abd Rahni Mt Piah, Universiti Sains Malaysia,

Malaysia

Image Authentication based on Faber Schauder DWT

Assma Azeroual, Karim Afdel

A Reflectance Transformation Imaing method for Large-scale objects

Yong Hwi Kim, Junho Choi, Yong Yi Lee, Bilal Ahmed, Kwan H. Lee

MLP Neural Network Classifier for Medical Image Segmentation

Manel JARRAR, Asma BEN ABDALLAH, Mohamed Hédi BEDOUI

< Sultan Moulay Sliman University ● Faculty of Sciences and Technics ● Room102>

Session cgiv2016_1.6: Visual Analytics Chair: Dr. Mohammed Abdulghafour, New York Institute of Technology , China

Evaluation of semantic similarity using vector space model based on textual corpus.

HSSINA badr, MERBOUHA Abdelkrim, BOUIKHALENE Belaid

Evolutionary algorithm for segmentation of medical images by region growing.

Ahmad El Allaoui, M’barek Nasri

Performance analysis of Streaming Video over Vehicular ad-hoc Network

ANSAM ENNACIRI, MOHAMED ERRITALI, MUSTAPHA MABROUKI, JAMAA BENGOURRAM

17:00 -

18:00


Chair: Ebad Banissi, London South Bank University, UK

CGiV2016_ Special Interest Group: Collaborative Research Meeting

19:30 -

21:30

cgiv2016 _ Social Networking Event

Social Networking Event Conference Dinner <further details shortly>

CGiV2016_PROGRAMME

5 | CGiV2016

Thursday 31 March 2016

09:00 < Sultan Moulay Sliman University ● Faculty of Sciences and Technics ● conference room hall>

Registration 09:30

– 11:15

< Sultan Moulay Sliman University ● Faculty of Sciences and

Technics ● Room102>

Session cgiv2016_2.1: Visualisation

Chair: Prof. Mohamed Fakir, Sultan Moulay Sliman University,

Morocco

Detecting abnormal events based on trajectory clustering

Najla Bouarada Ghrab, Emna Jammouci Fendri, Mohamed Hammami

Machine learning for real time poses classification using Kinect skeleton data

Youness CHOUBIK, Abdelhak Mahmoudi

Efficient Object Indexing and Retrieval Based on Partial Dominant Orientation Descriptor

Abdessamad Elboushaki, Rachida Hannane, Karim Afdel, Lahcen Koutti

Information Retrieval: A Comparative Study of Textual Indexing Using an Oriented Object Database (db4o) and the Inverted File

Mohamed ERRITALI

<short paper>

Diagnosis of Alzheimer Diseases in Early Step Using SVM (Support Vector Machine)

Amira Amira

A Comparative Study of Handwritten Mathematical Symbols Recognition

Yassine CHAJRI, Abdelkrim Maarir, Belaid BOUIKHALENE


Technics ● conference room>


Chair: Professor Muhammad Sarfraz, Department of Information

Science, Kuwait University, Kuwait

Roads Detection from Satellites Images Using a Region-based Active Contour Model and Distance Transform

Abdelkrim Maarir, Yassine Chajri, Belaid Bouikhalene

Features Selection based on Modified PSO Algorithm for 2D Face Recognition

Taher Khadhraoui, Sami Kktata, Faouzi Benzarti, Hamid Amiri

Spectral Graph Matching for Printed Tifinagh Character

Youssef OUADID, Brahim MINAOUI, Mohamed FAKIR

Face Recognition Based on City-block and Mahalanobis Cosine Distance

ABDELGHAFOUR ABBAD, KHALID ABBAD, HAMID TAIRI

Analysis of regional deformation of the heart's left ventricle using curvature values with Hotelling T2 metric

Rim Ayari

11:15 Morning Coffee Break

CGiV2016_PROGRAMME

6 | CGiV2016

11:45

- 13:15


Technics ● conference room >

Session cgiv2016_2.3: Image & Video Analysis

Chair: Ebad Banissi, LSBU, UK

A combined Approach Based on Fuzzy Classification and Contextual Region Growing to Image Segmentation

Mahaman Sani Chaibou, Karim Kalti, Mohamed Ali Mahjoub

Automatic Human segmentation in video using Convex Active Contours

Hiba Ramadan, Hamid Tairi

An effective foreground detection approach using a block-based background modeling

Omar Elharrouss, Driss Moujahid, Soukaina Elidrissi Elkaitouni, Hamid Tairi

Trabecular Bone Radiographs Characterization Using Lacunarity Measure

Ines SLIM, hanen Akkari, Asma Ben Abdallah, Imen Bhouri., Eric Lespessailles, Rachid Jennane, Mohamed Hedi Bedoui

Comparative study and enhancement of Camera Tampering Detection algorithms

Mohamed Ali Mahjoub, Mabrouka Hagui


Technics ● Room102 >


Chair: Prof. Mohamed Baslam, Sultan Moulay Sliman University,

Morocco

Ontology-based Approach for Online Arabic Character Recognition

Houda Nakkach

An effective Soft Multiple-Assignments strategies for enhancing the accuracy of the Content-Based Image Retrieval Systems

Zied ELLEUCH, Kirmene Marzouki

Detection and tracking of the moving objects in a video sequence by geodesic active contour

Mejda Chihaoui

New Approach of Visual Activity Measuring with Background Subtraction Algorithms

Kristian Dokic, Robert Idlbek, Antun Marinac

Pretreatment Approaches for Texture Image Segmentation

Khalid Salhi, El Miloud Jaara, Mohamed Talibi Alaoui

CGiV2016_PROGRAMME

7 | CGiV2016

11:45

- 13:15


Session cgiv2016_2.5: CGiV2016

Chair: Prof. Rachid El Ayachi, Sultan Moulay Sliman University, Morocco

<poster paper>

An Energy Consumption evaluation of Reactive and Proactive Routing Protocols in Mobile Ad-hoc Network

Mohamed Er-rouidi, Houda Moudni, Hicham Mouncif, Abdelkrim Merbouha

<short paper> Buildings change detection using multi-temporal satellite Quickbird imagery. Rabia Sarah Cheriguene, Habib Mahi

<short paper>

Semantic Similarity for Cross language plagiarism detection

HANANE EZZIKOURI, ERRITALI MOHAMMED, OUKESSOU MOHAMED

<short paper> Classification and recognition of dental images using a decisional tree Hicham RIRI, Abdelmajid ELMOUTAOUAKKIL, Abderrahim BENI HSSANE, Farid Bourezgui <short paper> Hierarchical Location-Based Services for Wireless Sensor Networks

RANIA KHADIM, MOHAMMED ERRITALI, ABDELHAKIM MAADEN


Lunch Break

CGiV2016_PROGRAMME

8 | CGiV2016

14:15

- 15:15


Technics ● conference Room>



Morocco

Face Classification under Different Kernel Function Compared to KNN Classifier

Bouchra Nassih, Aouatif Amine, Nabil Hmina

Local Dissimilarity Measures of Frames in Visual Substitution System for Blind People

Hanen Jabnoun, Frédéric Morain-Nicolier, Faouzi Benzarti, Hamid Amiri

Grouping K-means adjacent regions for semantic image annotation using Bayesian networks

Mustapha OUJAOURA, Rachid EL AYACHI, Brahim MINAOUI, Mohammed FAKIR, Omar BENCHAREF

Modelisation of raindrops based on declivity principle

Fouad ELHASSOUNI, Salma ALAMI, Abdelhak EZZINE


Technics ● Room102>


Chair: Prof. B. Bouikhalene, Sultan Moulay Sliman University,

Morocco

Hybrid Approach to Features Extraction for Online Arabic Character Recognition

Houda Nakkach

Comparative Study Between Iterative Estimator Parameters Of Hidden Markov Chain With Independent Noise for Segment Color And Gray Level Images

Meryem AMEUR, Najlae IDRISSI, Cherki DAOUI

SIFT descriptors modeling and application in texture image classification

Ossama ZEGLAZI, Aouatif AMINE, Mohammed RZIZA

Human identification based on the palmar surface of the hand Salma Ben Jemaa, Mohamed Hammami

14:15

- 15:15



Chair: Prof. Najlae Idrissi, Sultan Moulay Sliman University, Morocco

Facial Expressions Recognition using Decision Tree

Fatima zahra SALMAM, Abdellah MADANI, Mohamed KISSI

Automatic recognition of the damages and symptoms on plant leaves using parallel combination of two classifiers Ismail El Massi, Youssef Es-saady, Mostafa El Yassa, Abdeslam Benazoun, Driss Mammass

Channel Estimation and Equalization

Said ELKASSIMI, Said SAFI, Bouzid Manaut

On Performance Evaluation of Registration Algorithms for 3D Point Clouds

Mouna ATTIA, Yosr SLAMA, Mohamed Amine KAMOUN

CGiV2016_PROGRAMME

9 | CGiV2016

15:15 Afternoon Coffee Break 15:45

- 17:00


Technics ● Conference room>

Session cgiv2016_2.9: CGiV Application


Morocco

Difference Expansion Based Robust Reversible Watermarking with Region Filtering

Ka-Cheng Choi, Chi Man PUN

<poster paper>

Deterministic approach of visual servoing: robust object tracking by drone

Zakaria OUCHATTI, Alaa BENSAID, Fouad MOUTAOUAKKIL

Contribution to measure the user’s satisfaction of the public Moroccan libraries

JAOUAD OUKRICH, BELAID BOUIKHALENE

Pre-service teachers' perceptions and awareness toward Serious Games in the classroom – Case of Morocco

Ibrahim Ouahbi, Hassane Darhmaoui, Fatiha Kaddari, Abdelrhani Elachqar, Soufiane Lahmine


Technics ● Room 102> Session cgiv2016_2.10: Computer Aided Geometric Design

Chair: Prof. Cherki Daoui, Sultan Moulay Sliman University,

Morocco

A Novel approach for semi regular mesh based on planar proxies

Mejda Chihaoui, Akram Elkefi, Wajdi Bellil, Chokri Ben Amar

Finite element simulation of 2.5/3D shaped and rigid electronic circuits

Imen Chtioui, Frederick Bossuyt, Mohamed Hedi bedoui

Using Grammars in Medical Imaging: A Comprehensive Overview+

Salah Hamdi, Asma Ben Abdallah, Mohamed Hedi Bedoui

A blind identification chanel using multi-layer neural network

Elmostafa ATIFY, Ahmed Boumezzough, Cherki Daoui

O-LEACH of routing protocol for wireless sensor networks

wassim jerbi, abderrahmen guermazi, hafedh trabelsi

CGiV2016_PROGRAMME

10 | CGiV2016

15:45

- 17:00

< Sultan Moulay Sliman University ● Faculty of Sciences and Technics ● Room103> Session cgiv2016_2.11: Computer Aided Geometric Design

Chair: Prof. Mohamed Sabri, Sultan Moulay Sliman University, Morocco

Theoretical Improvement of the Image Compression Method Based on Wavelet Transform

Mourad Rahali, Habiba Loukil, Mohamed Salim Bouhlel

Recognition of 3D Objects using Heat Diffusion Equations and Random Forests

DRISS NAJI, FAKIR MOHAMAD, RACHID ELAYACHI, BELAID BOUIKHALENE

Recognition of Off-line Arabic Handwriting words Using HMM Toolkit (HTK)

Hicham El Moubtahij, Akram Halli, Khalid Satori

Optimization of the attribute vector by Genetic approach: Application to the classification of characters

Soukaina Benchaou, M'Barek Nasri, Fouad Aouinti, Khalid Zinedine, Ouafae El Melhaoui

17:00 -

17:30

< Sultan Moulay Sliman University ● Faculty of Sciences and Technics ● # > cgiv2017-Committee Members Meeting

19:15 -

Social & Cultural Networking Event

cgiv2016_Social & Cultural Networking Event Thursday 31 March 2016 – Social & Cultural Networking Event:

This tour has been organised for the conference delegates. Detail of this tour is as follows: Scheduled: Thursday 31st March 2016 Time: 19:15 – Starting:

Type of Function:

CGiV2016_PROGRAMME

11 | CGiV2016

Friday 1 April 2016 09:00 < Sultan Moulay Sliman University ● Faculty of Sciences and Technics ● Conference room hall>

Registration 9:30 < Sultan Moulay Sliman University ● Faculty of Sciences and

Technics ● Conference room> Session cgiv2016_3.1: MediViz – Medical Visualisation

Chair: Prof. Brahim Minaoui, Sultan Moulay Sliman University,

Morocco

Automatic Detection of Blood Vessel in Retinal Images

Abderrahmane ELBALAOUI, Mohamed FAKIR, Khaddouj TAIFI, Abdelkrim MERBOUHA

Grammar Formalism for Computed Tomography and Angiography Image Segmentation and 3D Carotid Artery Reconstruction


A Combined Approach for Breast Cancer Detection in Mammogram

Mohammed RMILI, Abdelmajid El Moutaouakkil, Abdellatif Siwane, Fatiha Essodegui, Fatiha Adnani

Enhancing EEG Surface Resolution by Using a Combination of Kalman Filter and Interpolation method

Ibtissem KHOUAJA BENFRADJ, Ibtihel NOUIRA, M. Hedi BEDOUI, Mohamed AKIL

Hemodynamic modeling in a stenosed internal carotid artery

Arij Debbich, Asma Ben Abdallah, Randa Salem, Patrick Clarysse, Mohamed Hédi Bedoui

Offline Chinese Handwriting Character Recognition through Extracting and Mapping Features to an Undirected Graphical Representation

Luo Yuechen, Xia Rui, Mohammed Abdulghafour


Technics ● Room102> Session cgiv2016_3.2: CGiV_Application

Chair: Prof. Mourad Nachaoui, Sultan Moulay Sliman University,

Morocco

Database of Handwritten Arabic Mathematical Formula Images Ibtissem Hadj Ali and Mohammed Ali Mahjoub

Efficient Video Summarization Based on Motion SIFT-Distribution Histogram

Rachida Hannane, Abdessamad Elboushaki, Karim Afdel

Face recognition using a fuzzy approach and a multi-agent system from video sequences

HICHAM HATIMI, MOHAMED FAKIR, MOHAMED CHABI

Image Analysis Using Disc-Harmonic Moments and Their RST Invariants in Pattern Recognition

Driss MOUJAHID, Omar ELHARROUSS, Hamid TAIRI

Improving pedestrian detection using Support Vector Regression

Mounir ERRAMI, Mohammed RZIZA

A comparison on performance of data mining algorithms in classification of whewellite and weddellite urinary human calculi images color using different descriptors

AIT IDER Abdelouahed, Driss. Naji, A. Ben Ali, S. Rabi M. Echajia, A. Merbouha and M. Mbarki

11:20 < Sultan Moulay Sliman University ● Faculty of Sciences and Technics ● Conference room hall>

Morning Coffee Break

CGiV2016_PROGRAMME

12 | CGiV2016

12:00 < Sultan Moulay Sliman University ● Faculty of Sciences and Technics ● Conference room>

Session cgiv2014_3.3: CGiV - Application

Chair: Prof. Pun, Chi Man, University of Macau, Macau S.A.R., China

Modeling of 2D objects with weighted-Quadratic Trigonometric Spline

Muhammad Sarfraz, Shamaila Samreen, Malik Hussain

Automatic Bifurcation Detection in Coronary X-Ray Angiographies

Asma Kerkeni, Asma BenAbdallah, Antoine Manzanera, Mohamed Hedi Bedoui

Conference Closing & Best Paper Awards from cgiv2016 committee:

Co-Chairman: Prof. Ebad Banissi, BDIRG, LSBU, UK

Prof. M. Fakir, Faculty of Sciences and Technics, Sultan Moulay Sliman University, Morocco

Prof. Ahamed Zeghal, Dean of Faculty of Sciences and Technics, Sultan Moulay Sliman University

Prof. Muhammad Sarfraz, Department of Information Science, Kuwait University, Kuwait 13:00 < Sultan Moulay Sliman University ● Faculty of Sciences and Technics ● Room101>

Lunch Break 13:30 Close

CGiV2016_Abstract

13 | CGiV2016

ABSTRACTS

Session CGiV2016_1.1: Visualisation Chair: Chair: Prof. Cherki Daoui, Sultan Moulay Sliman

University, Morocco

Cloud Motion Estimation in Satellite Image Sequences by Tracking Skeleton Critical Points using Lucas-Kanade method

Hassan Id Ben Idder, Nabil Laachfoubi

In this work we propose a new approach to estimate motion of clouds from satellite images using tools from digital geometry in combination with tools from computer vision field. The idea is to represent the cloud area by their binary skeleton and then to track the critical points of the pruned skeleton using optical flow estimation approach, particularly using the Lucas-Kanade Method which generates a sparse motion field of a set of feature points. The critical points of the skeleton can easily tell which pixel is suitable to be tracked by optical flow algorithm. Our method is motivated by the fact that critical points of skeleton can carry information about the global structure of an object. In the specific context of meteorological imagery, the information about the clouds shape and their topology is embedded in the skeleton critical points which makes them more suitable for determining the good feature to track by an optical flow algorithm. Keywords: Shape skeleton, Cloud motion estimation, Optical flow, Lucas-Kanade method, Satellite imagery

Fire Event Analytics: A Visual Analytics Tool to Assist Wildland Fire Management

Nikolaos Georgoulopoulos, Ioannis Mariolis, Savvas Rogotis, Dimos Ioannidis, Konstantinos Votis, Dimitrios Tzovaras

A comprehensive solution for early fire detection, prevention, monitoring, prediction and fire-fighting has been a major objective during the last

decades, due to numerous catastrophic wildland fires worldwide. In this work, Fire Event Analytics, a Visual Analytics tool that can operate as part of a fire event information system, is presented. Fire Event Analytics utilizes data acquired by unmanned aerial vehicles, effectively supporting wildland fire management. It is cloud-based and presents spatial and temporal cues to the end user, in order to form a comprehensive operations control module and enhance decision support. In that direction, fire growth prediction and simulation is supported, allowing hypothetical testing and comparison of multiple fire event scenarios, whereas a novel metric is introduced for assessing fire event criticality.

It consists of two internal components that present spatial and temporal cues to the user; Visual Analytics Module, and Decision Support Module. Visual Analytics Module is essentially the web interface of Fire Event Analytics. It consists of a main window presenting images from google maps API, and a control panel for data visualization and interaction with the map. The purpose of Decision Support Module (DSM) is to enhance the situation perception in the region of interest, so as to facilitate the end user to make optimal decisions for a possible fire event.

Fire Event Analytics supports various innovative functionalities; FARSITE integration into a web-based application, connection with weather APIs, use of digital elevation models and land-cover GeoTIFF data, hypothetical testing capability and comparisons of different fire growth simulations, support of multiple UAV surveillance and fire events, improvement of decision making with metrics (Chandler’s Burning Index, Fire Event Criticality Index) etc. Fire Event Analytics was tested with a fire event scenario of Israel Fire & Rescue Services, and Visual Analytics tool was able to cover the end users’ needs to increase situation awareness and support decision making. Keywords: fire management; decision support; visual analytics; UAV; fire detection

CGiV2016_Abstract

14 | CGiV2016

Session CGiV2014_1.2: Imaging Application Chair: Ahmad M. Aljamali, Kuwait University, Kuwait

Facial Feature Extraction and Recognition for Traditional Chinese Physiognomy

Maolin Huang, Jessica Liu

Chinese physiognomy is a traditional face reading method based on the inheritance(accumulate) of knowledge and experience. Through the face reading, one's personality, future, previous life and luck can be told. However, it is lack of the theory and systematic analysis to prove its outcomes.

This paper proposes a novel calculation method of personality based on the Chinese physiognomy. At first, we combined the ancient and the modem physiognomy to summarize the corresponding relation between the personality and facial feature and model the baseline to shape the face feature. Then the histogram of image is computed and the value of the histogram is employed by searching for the otsu threshold values to create a binary image in an adaptive way. The two-pass connected component method is indicated the feature region. Before the method, we erode the binary image remove the noise point, so that the new connected image can provide(achieve) a better result. Standard lines of hair and chin have been defined as the first line and the end line of the new connected images. A simple linear search method for eye area corner points, nose corner points and mouth corner points among the contours we have got. The method was tested on ORL face database. Keywords: Chinese physiognomy, personality, reading face

Robust Region Descriptors for Shape Classification

Cong Lin, Chi Man PUN

A novel scheme for efficient shape classification using region descriptors and extreme learning machine with kernels is proposed. The skeleton and boundary of the input shape image are first extracted. Then the boundary is simplified to remove noise and minor variations. Finally, region descriptors for the local skeleton, and the simplified shape signature are constructed to form a hybrid feature vector. Training and classification are

then performed using kernel extreme learning machine (k-ELM) for efficient shape classification. Experimental results show that the proposed scheme is very fast and can archive higher classification accuracy on the challenging MPEG-7 dataset, outperforming existing state-of-the-art methods. Keywords: Shape classification, Region Descriptor, Skeleton, Contour, Signature

Satellite Image Restoration by Applying the Genetic Approach to the Wiener Deconvolution

Fouad Aouinti, M'barek Nasri, Mimoun Moussaoui, Soukaina Benchaou, Khalid Zinedine

In the process of satellite imaging, the observed image is blurred by optical system and atmospheric effects and corrupted by additive noise. The image restoration method known as Wiener deconvolution intervenes to estimate from the degraded image an image as close as possible to the original image. The effectiveness of this method obviously depends on the regularization term which requires a priori knowledge of the power spectral density of the original image that is rarely, if ever, accessible, hence the estimation of approximate values can affect the restored image quality. In this paper, the idea consists of applying the genetic approach to the Wiener deconvolution for satellite image restoration through the optimization of this regularization term in order to achieve the best possible result. Keywords: Satellite image, Spatially invariant blur, Non-blind restoration, Wiener deconvolution, Genetic algorithm

Session cgiv2016_1.3: Computer Graphics Chair: Dr Dimitrios Tzovaras, Center for Research and

Technology Hellas Information Technologies Institute,

Greece

Local and Fast Fog Detection based on Saturation and RGB-Correlation around the Picture Vanshing Point

CGiV2016_Abstract

15 | CGiV2016

Salma ALAMI, Abdelhak EZZINE, Fouad ELHASSOUNI

Fog is a natural and meteorological phenomenon that seems to be very dangerous for road driving. In its presence, the driver has a high perturbation in his field of view and must redouble vigilance. Therefore, it’s primary to detect its presence to be able to adapt any advanced driver assistance system according to the density of fog. In this paper, we present a new local approach for detecting daytime fog based on the saturation and the correlation between the components of the RGB color space computed for a region around the picture vanishing point. Results show a good fog detection for synthetic pictures and fast time-computation. Keywords: Fog detection, Correlation, Driver assistant, Vanishing point

A stereo planar self-calibration method of a camera with variable intrinsic parameters

Ismail EL batteoui, Abderrahim Saaidi, Khalid Satori

In this work a robust technique of camera self-calibration with variable intrinsic parameters from unknown planar scene is presented. The main idea of our present method consists to use only two images of the scene by basing on the corresponding points between the two images which are numerous and easy to determine, the handling of only two images reduces remarkably the computation time of our method. From the projection of only two scene points and by using the image of absolute conic we will show that it’s possible to obtain a set of self-calibration equations according to the intrinsic camera parameters. These equations are non-linear which required formulating of a non-linear cost function according to the intrinsic camera parameters which its minimization permits us to estimate these parameters in each image. The obtained experimental results on the synthetic and real images show the robustness and the reliability of our method. Keywords: Self-calibration, Unknown planar scene, Variable intrinsic parameters, Non-linear cost function.

Surface Tension and Wettability Modeling for Flowing Liquids

Mariusz Zubrzycki, Jacek Raczkowski

The presented simulation model of surface tension and wettability based on physical properties of liquids is designed for use in computer graphics. Due to the relatively small surface tension forces the model is useful for simulating liquid of small volume such as droplets. This model can be used in conjunction with various fluid simulation methods, one of the most popular - Marker and Cell has been selected for this paper. The paper describes also a simple and rapid method of determining the liquid surface as a mesh of triangles. The presented method improves the final visual effect and is well suited for determining the surface of the droplets. The simulation method was applied to create realistic animations of flowing liquid droplets of different types. Keywords: animation, computational fluid dynamics, liquid simulation, liquid surface, natural phenomena

Attacks Against AODV Routing Protocol in Mobile Ad-Hoc Networks

Houda MOUDNI, Mohamed ER-ROUIDI, Hicham MOUNCIF, Benachir EL HADADI

A mobile ad hoc network is much more vulnerable to attack than a wired network due to the dynamic changes of the network topology, high mobility, limited physical security and lack of centralized administration. Unfortunately, the routing protocols are designed based on the assumption that all nodes trust each other and cooperate without maliciously disrupting the operation of routing. This paper analyzes the impact of security attacks on the performance of the AODV routing protocol. Simulations are setup in the NS-2 network simulator and the performance of the AODV routing protocol is discussed under black hole, flooding and rushing attacks. This analysis is provided in terms of performance metrics, such as a packet delivery ratio, the average end-to-end delay and normalized routing load. Our simulation based study shows that flooding attack drastically increases the routing overhead of the protocol. The black hole attack is deadly and severely affects the packet delivery ratio and brings up the average end-to-end delay to unacceptable

CGiV2016_Abstract

16 | CGiV2016

ranges. While the rushing attack has a less significant effect on the network performance.

Keywords: MANET, AODV, Black hole attack, Flooding attack, Rushing attack, Network Simulator (NS2)

Session cgiv2016_1.4: Imaging Chair: Dr. Assma Azeroual, Computer Systems and Vision

Laboratory, Morocco

Phase extraction method in fringe projection profilometry using Instantaneous frequency

Rachid IDRISSI EL-AZAMI, Benaissa BELLACH, Hamid TAIRI

Three-dimensional (3-D) surface profiles can be obtained by using non-contact optical profilometry techniques. Fringe projection profilometry is widely used for three dimensional (3-D) surface shape, measurement using phase-shifting (PS) methods with multiple projected fringe pattern or transform methods with single projected fringe pattern. In this paper, Our research has focused on the study of the phase extraction method from a single fringe pattern in fringe projection profilometry using the instantaneous frequency method as one of the relevant techniques in the field of signal processing. The main advantage of this method of analysis based on the notion of instantaneous frequency is that it requires only one fringe pattern, which made from this method well suited for real-time measurement. The effectiveness of the instantaneous frequency method is illustrated by their comparison with traditional methods. Keywords: profilometry, fringe pattern, phase extraction, instantaneous frequency

An Improvable Structure for Similarity Searching in Metric Spaces: Application on Image Databases

Youssef Hanyf, Hassan Silkan, Halima Labani

In last decades, the similarity search is very required in various fields such as pattern recognition, security, and multimedia databases. Although the

metric approach usefulness for speeding similarity search in complex databases, the searching cost optimization still an open problem. In this paper we propose an improvable pivot-based method which can improve its research efficiency based on the past users’ queries. Because images are the most data type which are concerned by the similarity search, the proposed method is tested on real images database. The experiments show that the proposed method can significantly improve its searching efficiency relying on queries resolution.

Keywords: similarity search, Image databases, Data indexing, content based image retrieval, metric access methods.

Face Recognition using Local Binary Probabilistic Pattern (LBPP) and 2D-DCT Frequency Decomposition

Dahmouni Abdellatif, Aharrane nabile, El Moutaouakil karim, Satori kalid

E-mail: [email protected]

Facial biometrics is an active modality that uses the face characteristics as argument of person identification. In this paper, we propose a new face recognition system basing on the Local Binary Probabilistic Pattern (LBPP) face representation and the global 2D-DCT frequency methods. The Local Binary Probabilistic Pattern is an alternative of the famous LBP descriptor which uses the confidence interval concept to evaluate the current pixel. Then the LBPP transformed images are decomposed in the frequency domain at 2D-DCT method to build a reduce features vector. The suggested approach is tested on ORL and Yale databases. The obtained results are very encouraging: 95.5% for ORL and 100% for Yale databases recognition rate.

Keywords: LBP, LBPP, Confidence interval, 2D-DCT

Session cgiv2016_1.5: Imaging Chair: Prof. Abd Rahni Mt Piah, Universiti Sains Malaysia,

Malaysia

mailto:[email protected]

CGiV2016_Abstract

17 | CGiV2016

Image Authentication based on Faber Schauder DWT

Assma Azeroual, Karim Afdel

The technology development has made the modification of digital content easier. The need of authenticating digital content is increasing. Image authentication can be done by embedding a mark in the image using digital fragile watermarking. In this paper we propose a new approach for image authentication based on Faber Schauder Discrete Wavelet Transform (FSDWT) and Singular Value Decomposition (SVD). The watermark used to authenticate the image is extracted from the image dominant blocks using SVD and FSDWT. This watermark is embedded in the LSB plan of image. This one is characterized by its contours and its around textures which contain an important concentration of dominant coefficients that are used to select the dominant blocks. Any image modification will result in an important change in the dominant blocks. Hence, if the image is altered, the singular values of the dominant blocks will be dramatically changed, then we confirm that the image is not authentic.

Keywords: Image authentication, Fragile watermarking, Tamper detection, Mixed scales FSDWT, Singular value Decomposition

A Reflectance Transformation Imaing method for Large-scale objects

Yong Hwi Kim, Junho Choi, Yong Yi Lee, Bilal Ahmed, Kwan H. Lee

RTI is an image-based rendering method which can represent the appearance of an object under varying illuminations. To create realistic synthetic-images using RTI, it is necessary to take dozens of images on a mounted camera with a calibrated point light source. Conventional RTI methods have proposed complex lighting systems in a hemispherical dome, or manually calibrate light poses using a reflective probe. In most cases, those methods are not suitable for the large-scale object in an outdoor environment because the size of the target object is restricted by the configuration of measurement systems.

In this paper, we present a new RTI method which can create photorealistic images of a large scale outdoor scene under arbitrary light directions. Instead of capturing RTI samples at a time for an entire

domain, we divide the RTI domain into a set of subsections. RTI samples in each section are acquired using a camera and an uncalibrated light source. After acquiring samples, we estimate svBRDF of measured samples without any prior knowledge of a 3D model, light poses, and surface normals. We also present an approach to merge the partial RTI images into a panoramic image. Experimental results show that our framework can extend RTI applicable to large-scale objects.

Keywords: light direction estimation, relighting image, reflectance transformation imaging, surface relief

MLP Neural Network Classifier for Medical Image Segmentation

Manel JARRAR, Asma BEN ABDALLAH, Mohamed Hédi BEDOUI

The choice of a segmentation method depends on several considerations, namely the nature of the image, the primitives to extract and the segmentation operations. We propose an MLP-basis neuronal approach for the choice of the segmentation method taking into account the nature of the input image. First, an evaluation of the quality of segmentation by different methods and using various criteria of evaluation was carried out. Then, a characterization of images, based on some objective parameters, was performed. The resulting descriptors will be used as input to the neuronal approach to associate each type of image with the adequate segmentation method after learning. We report the results of the intelligent choice of the segmentation method obtained on different databases of medical images. The discussion of these encouraging results allowed us to improve our success rate and cover all varieties of images.

Keywords: Segmentation, criteria of evaluation, descriptors, MLP-basis neuronal.

Session cgiv2016_1.6: Visual Analytics

Chair: Dr. Mohammed Abdulghafour, New York Institute of

Technology - China Branch, China

CGiV2016_Abstract

18 | CGiV2016

Evaluation of semantic similarity using vector space model based on textual corpus.

HSSINA badr, MERBOUHA Abdelkrim, BOUIKHALENE Belaid

In this work, we have created a semantic similarity calculation system between text documents to contribute to their semantic clustering. indeed, semantic clustering of documents is a promising field of research, since it guarantees a quick and targeted access to information. The aim of document clustering is to put together similar documents. We used the algebraic model VSM (Vector Space Model) [2] to represent text documents and the WordNet [1] lexical database, in that it groups words together based on their meanings. In this paper, we will present an overview of the static and semantic methods for calculating the similarity measure and the appropriateness of these methods.

As our research is focusing on the treatment of text documents on e-learning systems, we propose a corpus of a set of text documents. It consists of 32 text documents from the computer science textbook for high school students in Morocco. To evaluate our system, an experiment has been conducted among 298 students who produced 1192 text documents. Experimental evaluations using WordNet prove that the system presented in this work improves the accuracy of semantic similarity between the text documents.

Keywords: Corpus, E-learning, Semantic similarity, text document, WordNet.

Evolutionary algorithm for segmentation of medical images by region growing.

Ahmad El Allaoui, M’barek Nasri

Image segmentation by region growing method is robust fast and very easy to implemented, but it suffers from: the threshold problem, initialization, and sensitivity to noise. Evolutionary algorithms are particular methods for optimizing functions; they have a great ability to find the global optimum of a problem. In this paper, we used evolutionary algorithms to get over the three problems. We have proposed a

segmentation method based on region growing and evolutionary algorithms. The proposed approach is validated on four hundred synthetic images and medical. The results show the good performance of this approach.

Keywords: Segmentation, Region growing, Evolutionary algorithm.

Performance analysis of Streaming Video over Vehicular ad-hoc Network

ANSAM ENNACIRI, MOHAMED ERRITALI, MUSTAPHA MABROUKI, JAMAA BENGOURRAM

Nowadays Vehicular Ad hoc Networks is one of the most promising application areas of the wireless networks, able to organize without predefined infrastructure.

In particular, the transmission of video data in the VANETs represents a major issue due to the immense enrichment by the video information of the quality of experience perceived by the user and knowledge of the environment in many applications such as emergency video calls, road traffic conditions, tourist information, etc.

However, the video data transmission in VANETs is a difficult task, because of the stringent Quality of Service (QoS) requirements of video traffic. It then becomes necessary to establish specific transmission strategies to get over this problematic.

Accordingly, we are looking in this work about the quality of service QoS that allows the evaluation of the performance of a streaming transmission especially in real time. This concept of service quality will confront more constraints when it comes to ad-hoc networks .

At the same time we are interested to the Mobility analysis in vehicular ad hoc network and simulation of various performance factors in terms of bandwidth, loss rate, end-to-end packet delay, packet jitter and throughput of several vehicles nodes over the ad-hoc networks.

CGiV2016_Abstract

19 | CGiV2016

Our simulation is performed using the simulator OPNET Modeler in order to control the nature of multimedia traffic in our simualtions on different protocols.

Keywords: VANET, VIDEO STREAMING, QOS, OPNET

Session cgiv2016_2.1: Visualisation

Chair: Prof. Mohamed Fakir, Sultan Moulay Sliman University,

Morocco

Detecting abnormal events based on trajectory clustering

Najla Bouarada Ghrab, Emna Jammouci Fendri, Mohamed Hammami

Trajectories of moving objects provide crucial clues for video event analysis especially in surveillance applications.

In this paper, we proposed a novel approach for detecting abnormal events in video surveillance. Our approach is based on trajectory analysis involving two phases. In the first phase, we extracted clusters of normal events through an agglomerative hierarchical clustering of saved trajectories that were of different lengths, of different local time shifts and containing noise. Then, for each cluster a model was established. In the second phase, we aimed to classify a new event as normal or abnormal one. To achieve this objective, a comparison was performed with the extracted clusters' models thereby reducing the complexity and accelerating the classification process. Experiments were conducted to demonstrate the efficacy and the performance of our approach.

Keywords: Trajectories, abnormal, event detection, clustering, distance.

Machine learning for real time poses classification using Kinect skeleton

data

Youness CHOUBIK, Abdelhak Mahmoudi

Gestures recognition is an important research topic because some situations require silent communication (sign language gestures, surgeon gesture to the nurse for assistance… etc.). Traditionally, gesture recognition requires high quality expensive cameras and complicated computer vision algorithms. This is not the case thanks to the Microsoft Kinect sensor which provides an inexpensive and easy way for real-time user interaction. In this paper, we proposed a real time human gestures

CGiV2016_Abstract

20 | CGiV2016

classification technique, by using skeleton data provided by the Kinect sensor. Different users performed a set of tasks from a vocabulary of eighteen gestures. From skeleton data of each gesture, twenty features are extracted so that they are invariant with respect to the user’s size and its position in the scene. We then compared the generalization performances of four machine learning algorithms; support vectors machines (SVM), artificial neural networks (ANN), K-nearest neighbors (KNN) and Naive Bayes classifier (NB). Our method shows that SVM outperforms the other methods. Keywords: Gestures classification, Kinect sensor, SVM, Cross-validation

Efficient Object Indexing and Retrieval Based on Partial Dominant Orientation Descriptor

Abdessamad Elboushaki, Rachida Hannane, Karim Afdel, Lahcen Koutti

This paper describes a method for object indexing and retrieval using a new feature descriptor called Partial Dominant Orientation Descriptor (PDOD). The extraction process of the PDOD starts by sampling the object into a set of stable and informative key locations using Difference of Gaussian (DoG), so that the retrieval can proceed successfully despite changes in object viewpoint, scale, illumination, and distortion. The proposed descriptor at feature point takes into account the position and partially computes the dominant orientations of other key locations relative to this point, thus, offering a global distinctive and discriminative characterization. The extracted object descriptors are then indexed using Vocabulary Tree, which provides a robust object retrieval system across a substantial range of rotation variance, change in textures and colors, and object deformation. The extensive experiments on KONKLAB public dataset demonstrate that our method outperforms other benchmarks such as SIFT, PCA-SIFT and SURF indexing algorithms. Keywords: Object Indexing, Object Retrieval, Partial Dominant Orientation Descriptor (PDOD), Vocabulary Tree, Difference of Gaussian (DoG)

Information Retrieval: A Comparative Study of Textual Indexing Using an Oriented Object Database (db4o) and the Inverted File

Mohamed ERRITALI

The growth in the volume of text data such as books and articles in libraries for centuries has imposed to establish effective mechanisms to locate them. Early techniques such as abstraction, indexing and the use of classification categories have marked the birth of a new field of research called "Information Retrieval". Information Retrieval (IR) can be defined as the task of defining models and systems whose purpose is to facilitate access to a set of documents in electronic form (corpus) to allow a user to find the relevant ones for him, that is to say, the contents which matches with the information needs of the user.

Most of the models of information retrieval use a specific data structure to index a corpus which is called "inverted file" or "reverse index".

This inverted file collects information on all terms over the corpus documents specifying the identifiers of documents that contain the term in question, the frequency of each term in the documents of the corpus, the positions of the occurrences of the word...

In this paper we use an oriented object database (db4o) instead of the inverted file, that is to say, instead to search a term in the inverted file, we will search it in the db4o database.

The purpose of this work is to make a comparative study to see if the oriented object databases may be competing for the inverse index in terms of access speed and resource consumption using a large volume of data. Keywords: Information Retrieval, indexation, oriented object database (db4o), inverted file

<short paper>

Diagnosis of Alzheimer Diseases in Early Step Using SVM (Support Vector Machine)

Amira Amira

Alzheimer is a disease that affects the brain. It causes degeneration of nerve cells (neurons) and in particular cells involved in memory and intellectual functions. Early diagnosis of Alzheimer Diseases (AD) raises ethical questions, since there is, at present, no cure to offer to patients and medicines from therapeutic trials appear to slow the progression of

CGiV2016_Abstract

21 | CGiV2016

the disease as moderate, accompanying side effects sometimes severe. In this context, analysis of medical images became, for clinical applications, an essential tool because it provides effective assistance both at diagnosis therapeutic follow-up. Computer Assisted Diagnostic systems (CAD) is one of the possible solutions to efficiently manage these images. In our work; we proposed an application to detect Alzheimer’s diseases. For detecting the disease in early stage we used the three sections: frontal to extract the Hippocampus (H), Sagittal to analysis the Corpus Callosum (CC) and axial to work with the variation features of the Cortex(C). Our method of classification is based on Support Vector Machine (SVM). The proposed system yields a 90.66% accuracy in the early diagnosis of the AD.

Keywords: Alzheimer Diseases (AD); Computer Assisted Diagnostic(CAD) ; Hippocampus(H) ; Corpus Callosum (CC); Cortex(C ) ; Support Vector Machine (SVM).

A Comparative Study of Handwritten Mathematical Symbols Recognition

Yassine CHAJRI, Abdelkrim Maarir, Belaid BOUIKHALENE

Given the importance of mathematics in all branches of science (physics, engineering, medicine, economics, etc.), the recognition of handwritten mathematical expressions become a very important area of scientific research. The problem of recognizing mathematical expressions is divided into two sub-problems: the recognition of mathematical symbols and structural analysis of the mathematical expression.

In this research, we will focus on the recognition of handwritten mathematical symbols by the use of four descriptors for features extraction (GIST, PHOG, SURF and Centrist) and two classifiers (Artificial Neural Network (ANN) and Support Vector Machines (SVM)).

These experiments will be applied on a dataset which gathers the Arabic symbols and Latin symbols. This diversity of symbols and descriptors will allow us to make a comparison between these descriptors, classifiers and also a comparison between the Arabic and Latin symbols.

The paper is organized as follows: Section 2 presents some related works, section 3 describes the descriptors used in the features extraction phase and the last section analyzes and compares the results. Keywords: Mathematical symbols; recognition; features extraction; classification


Chair: Professor Muhammad Sarfraz, Department of Information

Science, Kuwait University, Kuwait

Roads Detection from Satellites Images Using a Region-based Active Contour Model and Distance Transform

Abdelkrim Maarir, Yassine Chajri, Belaid Bouikhalene

Automatic man-made objects detection from aerial and satellite images be a very important research field to understand the changes in our environment and gives an important source of information to be used in many fields. This study describes and evaluates an automated method for roads detection using a novel region-based active contour model and distance transform. The first step is pre-processing of images to reduce noses and increase the contrast between contours and to improve the quality of the initial image by using the median filtering, the second step deals with using a modified active contour with a new region-based signed pressure force (SPF) function that is adapted with the given images and in the last step centerlines of road are extracted with Euclidian distance transformation. The proposed method is tested on several images with high resolution and experiments results show that can detect both urban and suburban roads. Keywords: Roads detection, Active Contour, Distance transform, Image Segmentation

Features Selection based on Modified PSO Algorithm for 2D Face Recognition

CGiV2016_Abstract

22 | CGiV2016

Taher Khadhraoui, Sami Kktata, Faouzi Benzarti, Hamid Amiri

In this paper, we propose a Modified technique of features selection based on modified particle swarm optimization (MPSO) for face recognition system. PSO is a new class of algorithm for feature selection based on the idea of collaborative behavior of bird flocking. The feature selected by the proposed MPSO algorithm plays a vital role to search the solution space for an optimum solution where features are carefully selected according to a well defined discrimination criterion. Several novelties are introduced to make the recognition robust to varying illumination, facial expressions and poses at certain angles is challenging. Image of the face is divided first into sub-regions. Afterwards, the MPSO algorithm is applied to coefficients extracted by Discrete Wavelet Transform (DWT). We illustrate the experimental results of our new algorithm with the minimal set of selected features using different experimental protocols on several databases, including Yale Face, FEI and ORL.

Keywords: Biometrics, Face Recognition, Discrete Wavelet Transform, Particle Swarm Optimization, Feature Selection

Spectral Graph Matching for Printed Tifinagh Character

Youssef OUADID, Brahim MINAOUI, Mohamed FAKIR

Optical Character Recognition is one of the most important tools that contributes to facilitate man-machine interaction. In this paper, we present an optical Tifinagh character recognition system based on graph theory. After preprocessing, interest points are extracted using Harris corner detector. Based on these points we constructed the graph model representation of Tifinagh characters. Classification is done by calculating the spectral properties of adjacency matrix that represent the degree of agreement between graphs. The system shows satisfying performance and robustness against noise and deformation. The proposed system is evaluated using IRCAM database (Royal Institute of Amazigh Culture) and recognition rate of 99.02% was obtained.

Keywords: Character Recognition, Graph Matching, Spectral Method, Tifinagh Character, Image Character

Face Recognition Based on City-block and Mahalanobis Cosine Distance

ABDELGHAFOUR ABBAD, KHALID ABBAD, HAMID TAIRI

In this paper we propose a new powerful face recognition method to increase the performance of face recognition algorithms. In our idea we integrate two dissimilarity measures namely City-block and Mahalanobis Cosine distance. The experiments are performed on the ORL database and YALE database. The results indicate the interest of the proposed technique compared to others methods of literature.

Keywords: City-block distance, Facial biometrics, Mahalanobis Cosine distance, Measure of dissimilarity, Recognition rate

Analysis of regional deformation of the heart's left ventricle using curvature values with Hotelling T2 metric

Rim Ayari

This paper presents a robust method of local deformation analysis of the heart's left ventricle (LV) aimed at specifying the affected area. Our method is based on regional curvature variation calculation using the Hotelling T2 two samples difference metric . It has been validated with real data obtained from myocardial scintigraphy imaging techniques of 44 patients. For each patient we extract 3D surfaces representing the anatomical structures of interest, namely the endocardium and epicardium at stress. In order to progress to a regional analysis, we carried out, for each patient, a division into 17 regions according to the AHA standard.

Experimental results demonstrate the great robustness and efficiency of our method.

Keywords: Heart Left ventricle (LV), 3D object, AHA standard, Surface curvature, Hotelling T2 metric.

CGiV2016_Abstract

23 | CGiV2016


Chair: Ebad Banissi, LSBU, UK

A combined Approach Based on Fuzzy Classification and Contextual Region Growing to Image Segmentation

Mahaman Sani Chaibou, Karim Kalti, Mohamed Ali Mahjoub

We present in this paper an image segmentation approach that combines a fuzzy semantic region classification and a context based region-growing. Input image is first over-segmented. Then, prior domain knowledge is used to perform a fuzzy classification of these regions to provide a fuzzy semantic labeling. This allows the proposed approach to operate at high level instead of using low-level features and consequently to remedy to the problem of the semantic gap. Each over-segmented region is represented by a vector giving its corresponding membership degrees to the different thematic labels and the whole image is therefore represented by a Regions Partition Matrix. The segmentation is achieved on this matrix instead of the image pixels through two main phases: focusing and propagation. The focusing aims at selecting seeds regions from which information propagation will be performed. Thepropagation phase allows to spread toward others regions and using fuzzy contextual information the needed knowledge ensuring the semantic segmentation. An application of the proposed approach on mammograms shows promising results Keywords: Image Segmentation, Fuzzy Classification, Region-growing, Context Information, Contextual Region-growing

Automatic Human segmentation in video using Convex Active Contours

Hiba Ramadan, Hamid Tairi

This paper presents a new algorithm for full-automatic detection and segmentation of human in video. Our algorithm exploits the robustness and the accuracy of the interactive image segmentation using Convex Active Contours, to segment moving persons, but with an unsupervised manner. Based on a collaborative strategy to cluster a set of extracted

Selective Space Time Interest Points, the resulting separated moving clusters are used to initialize automatically the seeds for the segmentation. Experiments show a good performance of our algorithm for human detection and segmentation in video without a user interaction.

Keywords: human detection, segmentation, Convex Active Contours, collaborative clustering, Selective Space Time Interest Points

An effective foreground detection approach using a block-based background modeling

Omar Elharrouss, Driss Moujahid, Soukaina Elidrissi Elkaitouni, Hamid Tairi

The moving objects detection is considered as an important factor for many video surveillance applications. To assure a best detection a background model should be generated. This paper proposes a background modeling approach. To generate this model, we use both pixel-based and block-based processes to classify background pixels from those belong to the foreground. After that, to minimize the noise in the results of the background subtraction the structure-texture decomposition is applied on the absolute difference image. Just the structure component which contains the homogeneous parts of the image is used in the segmentation. The binary motion detection mask computation is made using a selected threshold. The experimental results demonstrate that our approach is effective and accurate for moving objects detection.

Keywords: Motion detection, Background subtraction, Background model, Video surveillance.

Trabecular Bone Radiographs Characterization Using Lacunarity Measure

Ines SLIM, hanen Akkari, Asma Ben Abdallah, Imen Bhouri., Eric Lespessailles, Rachid Jennane, Mohamed Hedi Bedoui

Osteoporosis is a disease characterized by low bone mass and deterioration of micro-architectural bone tissue, which provokes an

CGiV2016_Abstract

24 | CGiV2016

increased risk of fracture. This work treats the texture characterization of trabecular bone radiographs. The goal is to analyse according to clinical research a group of 174 subjects: 87 osteoporotic patients with various bone fracture types and 87 healthy subjects. In order to characterize osteoporosis, a method of lacunarity measurement, known as RDLC2, for grayscale image is used. This approach allowed the discrimination between healthy subjects and patients with osteoporosis. The results show an improved classification rate compared to another work [1]. Keywords: Texture analysis, Lacunarity, Osteoporosis

Comparative study and enhancement of Camera Tampering Detection algorithms

Mohamed Ali Mahjoub, Mabrouka Hagui

Camera Tampering Detection” uses several techniques based on image processing and computer vision. This study consists of the development of main camera tampering detection algorithms that can detect abnormal disturbance for video surveillance. Specially, a comparative study and enhancement of these algorithms was implemented. This developed system must take into account the various constraints and provides effective results.

Keywords: Camera Tampering detection; image processing; constraints; effective results.


Chair: Prof. Mohamed Erritali, Sultan Moulay Sliman University,

Morocco

Ontology-based Approach for Online Arabic Character Recognition

Houda Nakkach

Optical Character Recognition (OCR) is an active research area in pattern recognition. Since Arabic is among the most used language in the world. The exponential growth of use of new technologies, we are led to interest

in online script by proposing a new approach for character recognition using ontology methods. The main idea of this paper is bases on modelling Arabic characters by a proposed ontology. This ontology is composed by related sets of concepts. The concepts represent the strokes extracted from segmented characters. The objective behind this work is to study the role of ontology and their contributions in the recognition phase.

Keywords: ontology, strokes, online Arabic character recognition, spatial relation

An effective Soft Multiple-Assignments strategies for enhancing the accuracy of the Content-Based Image Retrieval Systems

Zied ELLEUCH, Kirmene Marzouki

The multiple-assignments approach alleviates the quantization error and enhances the accuracy of the Content-Based Image Retrieval (CBIR) systems. It aims to hard assign each feature vector to k-nearest visual words. However, during the matching step, the k-nearest visual words are used independently and ignore the significant of the best visual word. In this paper, we present our CBIR system which encapsulates several approaches such Hamming embedding, soft-assignment, multiple-assignments and graph fusion. We particularly focus on the multiple-assignments strategy. We propose an efficient soft multiple-assignments strategy to highlight the best k-nearest visual word. To this end, we explore the SOM topology which proved its performance in so doing. Moreover, we use graph fusion approach to fuse multi-features ranking lists.

Extensive experiments are conducted on Holiday and Ukbench public datasets. The experimental results are promising and outperform the state-of-the-art CBIR systems. In fact, we have reached KS score of 3.87 on Ukbench dataset and a mAP=85.6 on Holidays dataset.

Keywords: CBIR, soft-weighting, soft Multiple-Assignments, multi-features, SOM

CGiV2016_Abstract

25 | CGiV2016

Detection and tracking of the moving objects in a video sequence by geodesic active contour

Mejda Chihaoui

Detection of objects motion in image sequences is one of the most discussed research subjects in the artificial vision.

In this paper, we propose a new approach for the detection and tracking of one or more objects in video sequences.

Our proposed approach is based on geodesic active contour. In fact, the methods used in this domain generally rely on the detection by the geometric shapes that still have the isolated pixels constraints, i.e. they present a difficulty of extracting only the moving object and they do not generally allow detecting each object separately.

For this, our approach reduces these constraints by initially dividing the video into a succession of images. The active contour is, then, applied to each image till the end of the sequence. Finally, we reconstruct these new video images containing moving objects detected by the geodesic active contour.

Keywords: motion detection, tracking, video, geodesic active contour.

New Approach of Visual Activity Measuring with Background Subtraction Algorithms

Kristian Dokic, Robert Idlbek, Antun Marinac

Visual activity is a term that some authors use to describe motion and movement in video content and same authors also have proved increase in visual activity of video content during the last few decades. Background subtraction algorithms are commonly used for motion detection in the field of computer vision. In this paper one new approach for visual activity measuring with these algorithms will be discussed and compared with Visual Activity Index, the only available comprehensive approach for visual activity measuring. Keywords: visual activity, background subtraction, Visual Activity Index

Pretreatment Approaches for Texture Image Segmentation

Khalid Salhi, El Miloud Jaara, Mohamed Talibi Alaoui

In this paper, we present two approaches of the image texture pretreatment. The reason behind is to reduce the number of the grey level in the image, by assigning to each pixel a value that characterizes the local information of the neighborhood of this same pixel.

This coding process will allow us to reduce the size of the cooccurrence matrix and also minimize the extraction time of haralick features.

We compare these pretreatment approaches by applying them on our unsupervised segmentation method of the image texture, which is based on both Kohonen maps and mathematical morphology.

Our comparative study covers the results obtained by each pretreatment approaches taking into consideration the execution time and the error rate.

Keywords: Image Processing, Texture Image Clustering, Kohonen Network, Co-occurrence Matrix, Texture Pretreatment

Session cgiv2016_2.5: CGiV2016

Chair: Prof. Rachid El Ayachi, Sultan Moulay Sliman University,

Morocco

<short paper>

An Energy Consumption evaluation of Reactive and Proactive Routing Protocols in Mobile Ad-hoc Network

CGiV2016_Abstract

26 | CGiV2016

Mohamed Er-rouidi, Houda Moudni, Hicham Mouncif, Abdelkrim Merbouha

In mobile ad-hoc network (MANET) each node has the possibility to move freely in the space and communicate with each other over wireless link without any centralized controller or base station. These characteristics makes MANET useful and practical in several fields like military scenarios, sensor networks, Rescue operations, students on campus, etc. but this kind of network still suffers from a number of problems, power consumption is one of the most crucial design concerns in Mobile Ad-hoc networks as the nodes in MANET have battery limited. This paper will discuss about the power consumption aspect of the MANET routing protocols. A performance comparison of four routing protocols Dynamic Source Routing (DSR), Ad hoc On-Demand Distance Vector (AODV), Destination-Sequenced Distance Vector (DSDV) and Optimized Link State Routing (OLSR) with respect to average energy consumption are explained thoroughly. Then, an evaluation of how the varying parameters of network in diverse scenarios affect the power consumption in these four protocols is discussed. A detailed simulation model using Network Simulator 2 (NS2) with different mobility and traffic models is used to study their energy consumption.

Keywords: Ad-Hoc, Manet, AODV, DSR, DSDV, OLSR, Energy consumption

<short paper> Buildings change detection using multi-temporal satellite Quickbird imagery. Rabia Sarah Cheriguene, Habib Mahi

The Very High Spatial Resolution images (VHRS) is a powerful tool for quick mapping, especially in the detection of change occurring after a natural disaster. In this context, the aim of this paper, is to propose a methodology for detection of damaged buildings after an earthquake based on object-oriented classification applied to a pair of Quickbird images. The used dataset consists of three pan-sharpened channels with a spatial resolution of 0.6 m and size of 350 x 350 pixels acquired on the

city of Zemmouri (Algeria), before and after the earthquake of 6.8 magnitude occured on 21 May 2003. The adopted methodology is divided into three steps: In the first step, the multispectral images have been segmented using multiscale segmentation techniques with different values of the scale parameter. In this kind of segmentation technique, the homogenous regions are created taking into account the spectral features, spatial and textural features. Then the Borsotti criterion is used to select the best pair of segmentation results. In the second step, the nearest neighbor algorithm is applied in order to obtain an image of classes, the number of classes is fixed at 7 classes. After the process of classification each set of objects is assigned to a proper class and are connected with the class hierarchy. Knowing that the classification accuracy depends of the choice of the segmentation parameters such as the scale parameter and the fusion criterion.

The final step consists of extracting the buildings according to the value of rectangular descriptor which the value is less than 0.61.

The experimental results show that the detection rate was approximately 0.94 for existing buildings before earthquake and approximately 0.66 for damaged buildings. These rates are obtained by using the correctness parameters as quality measure. These results indicate the effectiveness of the proposed methodology

Keywords: change detection, object‐oriented classification, multiscale segmentation, very high spatial resolution <short paper>

Semantic Similarity for Cross language plagiarism detection

HANANE EZZIKOURI, ERRITALI MOHAMMED, OUKESSOU MOHAMED

Generally utterances in natural language are highly ambiguous, and a unique interpretation can usually be determined only by taking into account the context in the utterance occurred.Automatically determining the correct sense of a polysemous word is a complicated problem especially in multilingual corpuses.

CGiV2016_Abstract

27 | CGiV2016

This paper presents an application programming interface for several Semantic Relatedness/Similarity metricsmeasuring semantic similarity/distance between multilingual words and concepts, in order to use it after for sentences and paragraphs in Cross Language Plagiarism Detection (CLPD); using WordNet for the English-French and English-Arabic multilingual plagiarism cases. Keywords: Semantic Similarity, wordnet, Plagiarism Detection <short paper>

Classification and recognition of dental images using a decisional tree

Hicham RIRI, Abdelmajid ELMOUTAOUAKKIL, Abderrahim BENI HSSANE, Farid Bourezgui

Recognition and classification of images have a wide field of applications, especially in medical images. In order to provide to dentists a solution for classification of patient’s images to evaluate the evolution of their treatment, we need to use latest efficient technics of classification. In this paper, we propose an algorithm based on a decisional tree to classify and recognize 19 types of dental images. This hierarchical representation can be interpreted as a set of hierarchical types stored in leafs tree structure. By using several extracted features from color images acquired with a digital camera and grayscale images acquired by x-ray scanner. Such as facial features and skin color using YCbCr color space. The proposed technique has been evaluated on a large data set of four main types namely: mold, intra-oral, extra-oral and radiographic images of different patients. Hence, experimental results demonstrate the good performances of this approach. It shows that the average recognition rates can approaching of 100%.

Keywords: Recognition, classification, decisional tree, feature invariant, YCbCr, diagnosis aide, image analysis

<Poster paper> Hierarchical Location-Based Services for Wireless Sensor Networks RANIA KHADIM, MOHAMMED ERRITALI, ABDELHAKIM MAADEN

Nowadays Wireless Sensor Networks( WSNs) have attracted worldwide research and industrial interest, because they can be applied in various areas.

Geographic routing protocols for wireless sensor networks use Location information when they need to route packets. Obviously, location information is maintained by Location-Based Services provided by network nodes in a distributed way. The location based services is a distributed service without infrastructure in general. It has to answer to a location query such as: “Where is the node X”.

The location based services can be classified into two classes: Flooding-Based and Rendez-vous-based location services. We will take an interest on the second class specifically the hierarchical approach using the Hierarchical Location Service (HLS) and the Grid Location Service (GLS).

One pertinent performance criterion in the Location-Based Service is the network scalability. This work presents the scalability performance study of both GLS and HLS.Specifically, what happens if the number of nodes N increases. The study will focus on three qualitative metrics: The location maintenance cost, the location query cost and the storage cost. Also we compared using simulations two location based services, GLS and HLS while coupled to the well-known geographic routing protocol Greedy Perimeter Stateless Routing (GPSR) based on Localization performance. The simulations were performed using NS2 simulator for wireless sensor networks. Keywords: WSNs, Routing Protocols, Scalability, GLS, HLS, NS2 Simulator

session cgiv2016_2.6: Image & Video Analysis


Morocco

Face Classification under Different Kernel Function Compared to KNN Classifier

Bouchra Nassih, Aouatif Amine, Nabil Hmina

CGiV2016_Abstract

28 | CGiV2016

In this paper, we present a comparative study between Daubechies-DCT approach, Discrete Cosine Transform (DCT) and Histograms of Oriented Gradient (HOG) under different kind of kernel function. We obtain Daubechies-DCT by fusing the DCT features and Daubechies features. The implementation of HOG achieved by dividing the face image into small connected regions, named cells, and for each cell compiling a histogram of gradient directions. We use the fusion of DCT and Daubechies wavelets then HOG method to process face classification focused on SVM (Support Vector Machine) and KNN (K Nearest Neighbors) classifiers. The fusion features are inputted into SVM and KNN classifiers. Results show that the HOG with SVM-Rbf kernel function achieves the highest performance in terms of the detection rate which we obtained 96.5%. We present experimental results applied on MIT face database to demonstrate the effectiveness of a novel comparative study in terms of accuracy and running time.

Keywords: HOG, Daubechies wavelet, DCT, SVM, KNN

Local Dissimilarity Measures of Frames in Visual Substitution System for Blind People

Hanen Jabnoun, Frédéric Morain-Nicolier, Faouzi Benzarti, Hamid Amiri

As the environment identification is a vital necessity for blind people, a visual substitution system based on video analysis is the solution to their problem. This paper focuses on the assessment and integration of the local dissimilarity map in the video processing. A Real Value local Dissimilarity Map is built for grayscale images in order to get an excellent detection of similar frames. The elimination of similar frames processing allows reducing the redundancy of objects re-identification.

Keywords: Video processing, image comparison, Real Valued Local Dissimilarity Map, object recognition

Grouping K-means adjacent regions for semantic image annotation using Bayesian networks

Mustapha OUJAOURA, Rachid EL AYACHI, Brahim MINAOUI, Mohammed FAKIR, Omar BENCHAREF

To perform a semantic search on a large dataset of images, we need to be able to transform the visual content of images (colors, textures, shapes) into semantic information. This transformation, called image annotation, assigns a caption or keywords in a digital image. In this paper we try to resolve partially the region homogeneity problem in image annotation, we propose an approach to annotate image based on grouping adjacent regions, we use the k-means algorithm as the segmentation algorithm while the texture and GIST descriptors are used as features to represent image content. The Bayesian networks were been used as classifiers in order to find and allocate the appropriate keywords to this content. The experimental results were been obtained from the ETH-80 image database.

Keywords: Color; image; annotation; segmentation; descriptor; classification

Modelisation of raindrops based on declivity principle

Fouad ELHASSOUNI, Salma ALAMI, Abdelhak EZZINE

Performances of driver assistance systems depend on weather conditions: when the later deteriorate, for instance due to fog or rain, the performances deteriorate too. We hereby present a new approach of modelling raindrops on the windshield using the declivity principle. Our goal is to define a profile of raindrops by analyzing intensities behavior in different planes – horizontal, vertical, left-diagonal, and right-diagonal – based on the declivity principle’s properties.

Keywords: Raindrop detection declivity


CGiV2016_Abstract

29 | CGiV2016

Chair: Prof. Rachid El Ayachi, Sultan Moulay Sliman University,

Morocco

Hybrid Approach to Features Extraction for Online Arabic Character Recognition

Houda Nakkach

Recognition of Arabic Character field has been gaining more interest for many years, and a large number of research papers and reports have already been published in this area. There are several major issues with Arabic character recognition: Arabic characters are spelled differently (depending on whether they are isolated, at the beginning, in the middle or at the end of the word), multiple characters can have the same body but a number and/or position of various diacritics. The size of the Arabic characters may vary from one writer to another and even within the writing of a single writer, etc. This paper presents a new approach for feature extraction step of online handwritten Arabic character using global and local features. The system was tested with 2000 Characters written by different writers and the best rate of recognition obtained was 92.43%.

Keywords: Online Arabic Character Recognition, Feature extraction, Fourier descriptor, Chain Code

Comparative Study Between Iterative Estimator Parameters Of Hidden Markov Chain With Independent Noise for Segment Color And Gray Level Images

Meryem AMEUR, Najlae IDRISSI, Cherki DAOUI

The image segmentation is a fundamental tool to analyze and detect objects of interest that can be applied in many fields (medicine, satellite). In this work, we are going to present a classical Markov model for unsupervised image segmentation:

“Hidden Markov Chain with Independent Noise”(HMC-IN)

for segmenting both gray an color images. We compared five iterative algorithms EM, GEM, SEM, MCEM and ICE for estimating parameters of this model under two final bayesian decision criteria MAP and MPM

according to the execution time, the convergence, the PNSR index and the rate error.

Keywords: CE, EM, MCEM, GEM, SEM, MAP, MPM, Hidden Markov Chain with Independent Noise (HMC-IN), segmentation, iterative estimator, final Bayesian Decision Criteria.

SIFT descriptors modeling and application in texture image classification

Ossama ZEGLAZI, Aouatif AMINE, Mohammed RZIZA

This paper proposes a new statistical model for modeling the real images including texture images. Our model is based on the observation that The Scale-invariant feature transform (SIFT) descriptors extracted from an image can be properly modeled by the gamma distribution. The maximum-likehood estimation is used to solve for the two parameters of gamma distribution that gives the best fitting of the SIFT descriptors data. In addition, a new image feature based on gamma model is used to the texture image classification process. Experiments were conducted on brodatz database and demonstrate that our model leads to good improvements on term of the classification rates.

Keywords: Statistical image modeling, Gamma distribution, Classification, SIFT

Human identification based on the palmar surface of the hand

Salma Ben Jemaa, Mohamed Hammami

In this paper, we investigate a newmultimodal biometric approach for personal recognition, based on two different biometrics computed from the same hand image: palmprint

and fingerprint surface. The detection of these modalities is conducted with no constraints imposed on the subject. In addition, this approach has the particularity of using a new

fusion scheme based on the rank level integration method to consolidate the results obtained from the different biometric instances. Rank level fusion is a relatively new fusion approach. It is not a well-studied research

CGiV2016_Abstract

30 | CGiV2016

problem, yet it has a high potential for efficient consolidation of multiple unimodal biometric outputs matching. To the best of our knowledge, we present the first application of the rank level fusion to consolidate the rank information produced by the five individual fingerprint surface and the palmprint. The achieved experimental results have shown that the designed system achieves an excellent recognition rate and provides more security than unimodal biometric-based system. Keywords: Multimodal biometrics, Contactless hand detection, Palmprint recognition, Fingerprint surface recognition, Rank level fusion


Chair: Prof. Najlae Idrissi, Sultan Moulay Sliman University,

Morocco

Facial Expressions Recognition using Decision Tree

Fatima zahra SALMAM, Abdellah MADANI, Mohamed KISSI

Emotion recognition from facial expressions is generally performed in three steps: face detection, features extraction and classification of expressions. The present work focuses on two points: Firstly, a new extraction method is presented based on the geometric approach. This method consists of calculating six distances in order to measure parts of the face that better describe a facial expression. Secondly, an automatic supervised learning method called decision tree is applied on two databases (JAFEE and COHEN), in order to have a facial expressions classifying system with seven possible classes (six basic emotions plus neutral); this system uses as input the six distances previously calculated (using Euclidian ,Manhattan or Minkowski distance) for each face. Our results achieved a recognition rate of 89.20% and 90.61% respectively in JAFFE and COHEN database.

Keywords: Classification and Regression Tree, Emotions recognition, Facial expressions, Facial landmarks, Supervised Descent Method

Automatic recognition of the damages and symptoms on plant leaves using parallel combination of two classifiers Ismail El Massi, Youssef Es-saady, Mostafa El Yassa, Abdeslam Benazoun, Driss Mammass

The damages of pests and symptoms of parasitic diseases are among the major problems of agriculture in Morocco. They can cause important production and economic losses in the agricultural sector. The methods of pattern recognition and machine vision can be used in this context to reduce the losses. They allow providing decision support systems that will be used as mean of diagnosis and phytosanitary problem recognition from images. This study presents a multiple classifier system (MCS) for the identification of the damages and symptoms on plant leaves for a higher rate of recognition. The proposed approach is based on parallel combination of two kinds of classifiers including a neural network classifier that uses color, texture and shape features to discriminate between the classes, and an SVM classifier that uses texture and shape features. The decisions of two classifiers are combined using a combination module in order to obtain the final decision. Indeed, in order to elaborate our approach we have based on the methods used in three existing approaches in the field (Camargo et al. 2009; Al Bashish et al. 2011; Wang el al. 2012), in which the authors have adopted a single classifier in their systems. Firstly, we have implemented these approaches and we tested them on our database of images. Then, we have compared and analyzed the obtained results for designing our multiple classifier system. The tests of this study were carried out on six classes including three kinds of pest insects damages (Leaf miners, Thrips and Tuta absoluta), and three forms of fungal diseases symptoms (Early blight, Downy mildew and Powdery mildew). The experimental results show the efficiency of our approach compared to the previous existing approaches based on single classifiers, this approach is more effective and has the highest rate of recognition.

Keywords: Multiple classifier system, Parallel combination, Pattern recognition, Damage, Symptom, Image

Channel Estimation and Equalization

CGiV2016_Abstract

31 | CGiV2016

Said ELKASSIMI, Said SAFI, Bouzid Manaut

This paper present the problem of blind channel estimation of a non-minimum phase system using three algorithms. This algorithms play an important role for blindly channel estimation parameters. Thus studying the problem of blind equalization based on the presented algorithms. The simulation results in noisy environment and for different SNR demonstrate that the proposed algorithm is more powerful than other algorithms. In addition the proposed algorithm is more powerful in comparison to constant modulus algorithm (CMA) at the blind channel equalization.

Keywords: Adaptive blind equalization, Blind channel Identification, Higher Order Cumulants (HOC), Constant Modulus Algorithm (CMA), Bran A, BER, SER, Said1 Algorithm

Performance Evaluation of Registration Algorithms for 3D Point Clouds

Mouna ATTIA, Yosr SLAMA, Mohamed Amine KAMOUN

3D point Geometric alignment is a challenging task encountered in many scientific applications related to different fields such as robotics and computer vision. For this reason, the well-known 3D registration problem has been extensively studied, and a lot of efficient 3D registration algorithms (RA) exist. Even though many surveys in the litterature addressed RA’s, none to our knowledge is especially interested in their use in robotic fields and more precisely in dimensional control of mechanical pieces. Our present work involving both a theoretical and an experimental study compares some local and non-rigid RAs, used to align large point clouds representing mechanical pieces. This paper is two-fold and permits first to uncover the similarities and differences between four known RAs which are ICP, NDT, Softassign and RANSAC and then to establish an inter RAs comparative performance evaluation based on accuracy, speed and other new specific metrics we have defiened.

Keywords: 3D registration algorithm, accuracy, dimensional control, robotics, speed.

Session cgiv2016_2.9: CGiV Application

Chair: Prof. Mohamed Gouskir, Sultan Moulay Sliman University,

Morocco

Difference Expansion Based Robust Reversible Watermarking with Region Filtering

Ka-Cheng Choi, Chi Man PUN

Existing robust reversible watermarking methods usually have poor visual quality or unstable robustness and reversibility, implies that reversibility cannot be assured even in lossless channel. In this paper, a novel robust reversible watermarking method is proposed. In the proposed method, bit plane manipulation is applied to hide watermark bits in bit planes that are lesser affected by attacks. Region filtering is also adopted to find blocks that result in low variance for watermark embedding to further increase its robustness. Experimental results show that our method has improved performances compared with state-of-the-art technology, better surviving bit rate, a robustness measure, is obtained in the proposed algorithm. Keywords: bit plane; lossless data hiding; threshold; robust; reversible watermarking. <poster paper>

Deterministic approach of visual servoing: robust object tracking by drone

Zakaria OUCHATTI, Alaa BENSAID, Fouad MOUTAOUAKKIL

This work is part of an application context, focused on the analysis of flow of images acquired by a camera embedded in a drone, controlled by a control station. Specifically, we are interested in the coupling vision / command, to develop a control system that allows an autonomous navigation and operation of the Unmanned Aerial Vehicle in complex environments where the use of visual sensors appears to be essential for moving the drone in a controlled manner but also be capable of increasing the stability of the UAV.

CGiV2016_Abstract

32 | CGiV2016

We propose the state of the art in image-based visual servoing, focused on the control of unmanned aerial vehicles and allows to move from a current position to a desired position according the observed scene while also improving the flight performance (stability and accuracy).

Our contribution will be in the proposal of a control approach capable to tracking an object by introducing the concept of real-time and subsequently uses a visual memory in the form of keyframes to automatically reproduce a route already made.

Keywords: visual servoing, object tracking, motion estimation, drone

Contribution to measure the user’s satisfaction of the public Moroccan libraries

JAOUAD OUKRICH, BELAID BOUIKHALENE

In recent years, the number of students and documents resources has grown rapidly, which means that academic institutions are obliged to attach a great importance to the management of these resources. In this study, we conducted a survey method to lead this management and to investigate users’ information needs and library usage preference, the case of Sultan Moulay Slimane University libraries of Beni Mellal.

The main objective of this paper is to know the students’ information needs and to identify the performance of libraries services by using the descriptive analysis and Pareto diagram.

Keywords: Document, descriptive analysis, libraries services, survey, students

Pre-service teachers' perceptions and awareness toward Serious Games in the classroom – Case of Morocco

Ibrahim Ouahbi, Hassane Darhmaoui, Fatiha Kaddari, Abdelrhani Elachqar, Soufiane Lahmine

We conducted a study on the perception towards videogames among pre-service teachers in three Regional Teachers' Training Centers in Morocco (Centres Régionaux des Métiers de l'Education et de la Formation

(CRMEF)). 180 pre-service teachers responded to our primary general survey which mainly focused on their perceptions and attitudes towards the integration of video games in the classroom. We then conducted two training and sensitization sessions with 40 of these trainees on how to make and integrate games in the classroom. We centered our training sessions on the use of the SCRATCH tool which is an open source environment facilitating videogame creation. At the end of these training sessions, we conducted another survey to measure the evolution of their perceptions towards educational video games. This comparative study showed a considerable progress and a better awareness of these pre-service teachers about the potential of serious games in the classroom. All participants appreciated Scratch and its potential as a learning environment. They highlighted the importance of including serious games in class and their need for extra training on the subject. Two of these pre-service teachers, lately adopted this innovative approach and implemented it in their final projects. They have developed and successfully experimented one mathematics lesson using the serious game “Timez ATTACK” with grade three students. Keywords: Serious game, pre-service teachers, teachers training, Scratch environment, perception study

Session cgiv2016_2.10: Computer Aided Geometric Design

Chair: Prof. Mostapha Jourhmane, Sultan Moulay Sliman

University, Morocco

A Novel approach for semi regular mesh based on planar proxies

Mejda Chihaoui, Akram Elkefi, Wajdi Bellil, Chokri Ben Amar

Abstract: Surface meshes have recently attracted great interest since they are frequently used in many computer graphics applications. These meshes are often generated by isosurfacing representations or also by scanning devices. Unfortunately, such meshes are dense and they contain many redundant vertices and irregular sampling. This make meshes unable to support multiple applications as storage, display,

CGiV2016_Abstract

33 | CGiV2016

compression or transmission. To solve these problems and reduce the complexity, the mesh quality (connectivity regularity) must be improved. This improvement is called remeshing. On the other hand, many researches techniques have been done to exploit a mesh' local planarity to optimize its geometric representation. This paper presents a novel multiresolution approach of remeshing based on planar proxies.

Keywords: surface meshes, remeshing, semi-regular mesh, multiresolution, approximation, geometric primitive, shape proxies

Finite element simulation of 2.5/3D shaped and rigid electronic circuits

Imen Chtioui, Frederick Bossuyt, Mohamed Hedi bedoui

Today a need is emerging for embedding electronic and sensor functions in the products which needs these functions, and, importantly, to do this without noticeably influencing the mechanical design of the product. This contribution describes an approach used to produce a 2.5/3D free-form rigid and smart objects or shaped circuit.

The proposed fabrication process of shaped circuit is compatible with a typical printed circuit manufacturing and electronics assembly. Once the circuit is completed in its flat shape,its random final functional shape is given using thermoforming. In order to be able to deform a given flat circuit to its final form with predictable final spatial positions of components and interconnections. a FEM simulation is conducted to model the thermoforming of polymer based electronics circuit.

As one of the process outputs, the wall thickness distribution predicted for the final part is compared with the experimental results . Keywords: 2.5/3D circuit, Thermoforming, Numerical simulation, Hyperelastic

Using Grammars in Medical Imaging: A Comprehensive Overview


The aim of this article is to present a comprehensive overview concerning the use of syntactic pattern recognition methods in medical image analysis and representations in quite a lot of applications. To complete this purpose, we used a systematic procedure. The used methods and techniques were extorted from each study, the primitives used, the results obtained in medical images using grammars, and techniques used for learning the grammatical rules. Grammars can represent different lesions in medical images such as cardiac and scintigraphic images, coronaries arteries and bone radiographs. Grammar formalisms proved to be useful for medical data processing. The reviewed methods and techniques allowed a comprehensive overview. Most of articles were published in the last decade. Furthermore, it can be concluded that it is an extremely promising research line, especially when the studies showed a recognition rate more than 90%.

Keywords: Medical imaging, Computer vision, Pattern recognition, Grammars, Syntactic methods

A blind identification chanel using multi-layer neural network

Elmostafa ATIFY, Ahmed Boumezzough, Cherki Daoui

In this paper we try to propose an approach using multilayer neural networks to model the blind identification method. This method is often used in the case of identification supervise nonlinear channels. In our work we propose equations which allow adaptation of the network to identify the transmission channel by a MC-CDMA supervise approach.

Keywords: blind identification, equalization, neural network, adaptive filter, FIR channel

O-LEACH of routing protocol for wireless sensor networks

wassim jerbi, abderrahmen guermazi, hafedh trabelsi

LEACH protocol called Low Energy Adaptive Clustering Hierarchy, is a protocol that allows the formation of distributed cluster. In each cluster, LEACH randomly selects some sensor nodes called cluster heads (CHs).

CGiV2016_Abstract

34 | CGiV2016

The selection of CHs is made with a probabilistic calculation. It is supposed that each non-CH node joins a cluster and becomes a cluster member. Nevertheless, some CHs can be concentrated in a specific part of the network. Thus several sensor nodes cannot reach any CH. to solve this problem. We created an O-LEACH Orphan nodes protocol, its role is to reduce the sensor nodes which do not belong the cluster. The cluster member called Gateway receives messages from neighboring orphan nodes. The gateway informs CH having the neighboring nodes that not belong to any group. However, Gateway called (CH') attaches the orphaned nodes to the cluster and then collected the data. O-Leach enables the formation of a new method of cluster, leads to a long life and minimal energy consumption. In orphan node possess enough energy and seeks to be covered by the network. The principal novel contribution of the proposed work is O-LEACH protocol which provides coverage of the whole network with a minimum number of orphaned nodes and has a very high connectivity rates.

As a result, the WSN application receives data from the entire network including orphan nodes. The proper functioning of the Application supported therefore requires management of intelligent resources present within each the network sensor. The simulation results show that O-LEACH performs better than LEACH in terms of coverage, connectivity rate, energy and scalability. Keywords: WSNs; routing; LEACH; O-LEACH; Orphan nodes; sub-cluster; gateway; CH’.

Session cgiv2016_2.11: Computer Aided Geometric Design

Chair: Prof. Mohamed Sabri, Sultan Moulay Sliman University,

Morocco

Theoretical Improvement of the Image Compression Method Based on Wavelet Transform

Mourad Rahali, Habiba Loukil, Mohamed Salim Bouhlel

Image compression was performed by several techniques for example: JPEG and JPEG2000 are lossy compression methods. These methods

performing scalar quantization on the values obtained after transformation. The disadvantage of the scalar quantization is it does not allow exploiting the spatial correlation between pixels in the image. To improve the compression, we quantified together of values simultaneously it is definition of the vector quantization. In this paper, we studied and modeled an approach to images compression by wavelet transform and Kohonen network. We show the role of null moments in wavelet for improve the compression and we calculate the compression ratio based on compression parameters. Keywords: Compression ratio, moments null, neural network, wavelet transform

Recognition of 3D Objects using Heat Diffusion Equations and Random Forests

DRISS NAJI, FAKIR MOHAMAD, RACHID ELAYACHI, BELAID BOUIKHALENE

In this paper, we present an approach to recognize 3D objects. The method is based on the heat equation by calculating the geodesic distance between any pair of points x, y on Riemannian manifold. The method is compared to the light field descriptor (LFD) and the salient views (SV) by using three classifier such as C4.5 decision tree and Random Forest. The proposed method is a set of classifier that makes the decision, referring to the votes. We evaluate the performance of the proposed approach on a very challenging task of recognizing different objects types using two databases.

Keywords: 3D object, Recognition, Heat diffusion, Geodesic distance, Random Forests. Recognition rate.

Recognition of Off-line Arabic Handwriting words Using HMM Toolkit (HTK)

hicham el moubtahij, Akram Halli, Khalid Satori

There are a lot of difficulties facing a good handwritten Arabic recognition system such as the similarities of different character shapes and the

CGiV2016_Abstract

35 | CGiV2016

unlimited variants in human handwriting. This paper presents a handwriting Arabic word recognition system. The objective of this approach is to propose an analytical offline recognition method of handwritten Arabic for rapid implementation.

The first part in the writing recognition system is the preprocessing phase that prepares the data which serves to introduce and extract a set of simple statistical features by a window sliding along that text line from the right to left; then it injects the resulting feature vectors to the Hidden Markov Model Toolkit (HTK). In the recognition phase, the concatenation of characters to form words is modelled by simple lexical models; each word is modelled by a stochastic finite-state automaton (SFSA). The proposed system is applied to an “Arabic-Numbers” data corpus, which contains 47 words and 1905 sentences. These sentences are written by five different peoples.

Keywords: Arabic text handwritten, Hidden Markov Model Toolkit (HTK), stochastic finite-state automaton

Optimization of the attribute vector by Genetic approach: Application to the classification of characters

Soukaina Benchaou, M'Barek Nasri, Fouad Aouinti, Khalid Zinedine, Ouafae El Melhaoui

E-mail: [email protected]

Selecting the parameters for the classification is a delicate process. We present in this paper a method for selecting the parameters by the genetic algorithm which optimizes the choice of parameters by minimizing a cost function. This function is defined by a Trace criterion. The approach is validated on some characters images. The proposed algorithm gives a fast convergence towards the optimal solution.

Keywords: Genetic algorithm, selection of attributes, image processing, characters recognition, classification methods.

Session cgiv2016_3.1: MediViz – Medical Visualisation

Chair: Prof. Brahim Minaoui, Sultan Moulay Sliman University,

Morocco

Automatic Detection of Blood Vessel in Retinal Images

Abderrahmane ELBALAOUI, Mohamed FAKIR, Khaddouj TAIFI, Abdelkrim MERBOUHA

Automatic detection of retinal blood vessels and measurement of vessel diameter are very much important for the diagnosis and the treatment of different ocular diseases including diabetic retinopathy (DR), glaucoma and hypertension. In this paper, we present a novel method to detect blood vessels in the fundus retinal images. The proposed method consists of two main steps. The first step is pre-processing of retinal image to improve the retinal images by evaluation of several image enhancement techniques. In the second step, the vesselness filter is usually used to enhance the blood vessels and Hessian multiscale enhancement filter is designed from the orientation histogram of the output of the vesselness filter. The performance of algorithms is compared and analyzed on three publicly available databases (DRIVE, STARE and CHASE_DB1 ) of retinal images using a number of measures, which include accuracy, sensitivity, specificity and operating characteristic (ROC) curve. The results (DRIVE: Se = 0.76, Sp = 0.97; STARE: Se = 0.77, Sp = 0.97. CHASE_DB1 : Se = 0.75, Sp = 0.95) and comparison with other methods show the high accuracy of the proposed method.

Keywords: Blood Vessel Detection, Vesselness filter, Hessian multiscale enhancement filter

Grammar Formalism for Computed Tomography and Angiography Image Segmentation and 3D Carotid Artery Reconstruction


Understanding the medical image content has always been a central problem in the area of computer vision and pattern recognition, and a possible approach to solve this problem is the use of syntactic methods. Grammars can represent different lesions in medical images cardiac and scintigraphic images, coronaries arteries and bone radiographs. This

mailto:[email protected]

CGiV2016_Abstract

36 | CGiV2016

paper shows that grammar formalisms proved to be useful for Computed Tomography and Angiography (CTA) image. Practically a set of 80 real CTA images have been segmented based on plex grammar. The experiment results allow making 3D carotid artery reconstruction; this work is intended for medical diagnosis assistance and computer-aided diagnosis systems. Keywords: CTA image; Segmentation; Grammar formalism; 3D carotid artery reconstruction

A Combined Approach for Breast Cancer Detection in Mammogram

Mohammed RMILI, Abdelmajid El Moutaouakkil, Abdellatif Siwane, Fatiha Essodegui, Fatiha Adnani

Mammography is the best screening tool used today to find breast cancer. The method for Image Segmentation Mammographic proposed in this paper is based in one hand on Irregular pyramidal architecture, and in the other hand on an algorithm using together a k-means and Multi-step Maxmin. The original image is fragmented (quadtree) initially to homogeneous regions. Each region is then associated with a peak of graph and we use topographical distance between two regions. It gathers data within homogeneous groups named regions classes, then we use our algorithm to find the optimal partition of regions which have the minimum value of entropy. This technique gives good results, and allows calculating morphological parameters of the breast cancer.

Keywords: Mamography, Image Segmentation, K-means, Irregular pyramid, Multi-step Maxmin, topographical distance.

Enhancing EEG Surface Resolution by Using a Combination of Kalman Filter and Interpolation method

Ibtissem KHOUAJA BENFRADJ, Ibtihel NOUIRA, M. Hedi BEDOUI, Mohamed AKIL

With recent progress in the medical signals processing, the EEG allows to study the Brain functioning with a high temporal and spatial resolution. This approach is possible by combining the standard processing algorithms of cortical brain waves with characterization and interpolation

methods. First, a new vector of characteristics for each EEG channel was introduced using the Extended Kalman filter (EKF). Next, the spherical spline interpolation technique was applied in order to rebuild other vectors corresponding to virtual electrodes. The temporal variation of these vectors was restored by applying the EKF. Finally, the accuracy of the method has been estimated by calculating the error between the actual and interpolated signal after passing by the characterization method with the Root Mean Square Error algorithm (RMSE).

Keywords: EEG, Extended Kalman Filter, Signal characterization, Spherical Spline Interpolation, Root Mean Square Error

Hemodynamic modeling in a stenosed internal carotid artery

Arij Debbich, Asma Ben Abdallah, Randa Salem, Patrick Clarysse, Mohamed Hédi Bedoui

This paper describes the blood flow modeling in a patient-specific carotid artery bifurcation suffering from an internal carotid stenosis. This disease can hamper the brain irrigation and even cause a stroke. Our aim is to predict the hemodynamic behavior through computational fluid dynamic study. Our approach realizes a hemodynamic modeling within a geometric carotid model build from a 3D computed tomography angiography image with blood considered as a Newtonian and incompressible fluid, and the wall as rigid. The blood flow modeling is based on the Navier-Stokes equation. A Womersley velocity profile is used as a boundary condition in the common carotid artery. Main results of this study are the following: velocity is maximum in stenosis and minimum in the sinus. Pressure is negative on the most of carotid artery bifurcation unless the post-stenotic site.

Keywords: Hemodynamic modeling, carotid artery bifurcation, blood flow, atherosclerosis, CFD

Offline Chinese Handwriting Character Recognition through Extracting and Mapping Features to an Undirected Graphical Representation

CGiV2016_Abstract

37 | CGiV2016

Luo Yuechen, Xia Rui, Mohammed Abdulghafour

In this paper, new methods were developed to successfully identify Chinese handwriting characters. These methods are based on features extraction as compared and matched with HCL2000 database [1]. Several algorithms were applied for binarization, smoothing, noise reduction and thinning to an image of a single Chinese character. Then the image is given to a structural feature extracting algorithm, which transforms that character into an undirected graph with unique coordinates of all nodes. The resulting graph was compared with 3755 samples from the database one by one, whose features are also extracted and stored in a graph. The total deviation between two characters was obtained by comparing edges from the generated undirected graph representing a character and its counterpart graph which is generated from the image in HCL 2000 database. Based on the measurements of lengths, orientation, and areas between lines, the best match was selected as the result of recognition. Additional principles are also included in order to assure the accuracy. During the matching, the graph may sometimes be slightly transformed or modified to maximize the fitness criteria. Experimental results are accomplished by the use of 1000 random characters to test the effectiveness. The accuracy of the recognition system is significant. Analysis and experimental results are presented.

Keywords: Feature extraction, mapping, matching, undirected graph, Chinese character recognition, HCL 2000

Session cgiv2016_3.2: CGiV_Application

Chair: Prof. Mourad Nachaoui, Sultan Moulay Sliman University,

Morocco

Database of Handwritten Arabic Mathematical Formula Images Ibtissem Hadj Ali and Mohammed Ali Mahjoub

Although publicly available, ground-truthed database have proven useful for training, evaluating, and comparing recognition systems in many domains, the availability of such database for handwritten Arabic

mathematical formula recognition in particular, is currently quite poor. In this paper, we present a new public database that contains mathematical expressions available in their off-line handwritten form. Here, we describe the different steps that allowed us to acquire this database, from the creation of the mathematical expression corpora to the transcription of the collected data. Currently, the dataset contains 4 238 off-line handwritten mathematical expressions written by 66 writers and 20 300 handwritten isolated symbol images. The ground truth is also provided for the handwritten expressions as XML files with the number of symbols, and the MATHML structure.

Keywords: Mathematical expression recognition; database; Handwritten; Arabic formula.

Efficient Video Summarization Based on Motion SIFT-Distribution Histogram

Rachida Hannane, Abdessamad Elboushaki, Karim Afdel

Video summarization refers to the process of recapitulating video stream by producing an abstract of the salient keyframes that could cover its overall content. However, an efficient video summarization requires an efficient video shot boundary detection (SBD) and keyframes extraction. In this backdrop, this paper presents a novel and efficient approach for video SBD and keyframes extraction that will lead in the summarizing video. Meanly, Motion SIFT-Distribution Histogram (MoSIFT-DH) is extracted from the frames as a Glocal feature. The shot boundaries are detected using an adaptive threshold for the computed distance of MoSIFT-DH of the consecutive frames. Furthermore, keyframe representing the salient content of each segmented shot is extracted using entropy based singular values. Finally, the summarizing video is generated, by combining all the extracted keyframes. Our experiments on various videos indicate that our method can efficiently detect shot boundaries under different levels of illumination, camera operations and motion effects.

CGiV2016_Abstract

38 | CGiV2016

Keywords: Video Summarization, Shot Boundary, Keyframe Extraction, Motion SIFT

Face recognition using a fuzzy approach and a multi-agent system from video sequences

HICHAM HATIMI, MOHAMED FAKIR, MOHAMED CHABI

Face recognition systems in a video sequence constitute essential technical tools in several domains. To classify the faces in minimal time, the classic methods of classification being inadequate, fuzzy logic is considered as an effective technique for solving a classification problem. This article proposes a fuzzy approach for detection and face recognition in video sequences using a multi-agent modeling. This method contains several steps to classify the faces detected in the video. The multi-agent approach adopted allows minimizing the complexity of the processing and getting to the result with minimal time. The tasks of detection and classification of face are realized in two steps. In the first step, faces are detected using texture color and geometrical face. In the second step, the multi-agent system and fuzzy approach are used in the recognition process to find the degrees of membership. The results obtained using this method demonstrates performance in terms of robustness, in the variations illumination and speed.

Keywords: Fuzzy classifier, Fuzzy inference, face Recognition, Face detection, Multi-agent

Image Analysis Using Disc-Harmonic Moments and Their RST Invariants in Pattern Recognition

Driss MOUJAHID, Omar ELHARROUSS, Hamid TAIRI

Moments and moment invariants are the most useful tools in pattern recognition. Recently, the Conventional Disc-Harmonic Moments (CDHMs) are used to describe binary and gray scale images. In order to deal with color images in a holistic manner, these CDHMs are generalized as Quaternion Disc-Harmonic Moments (QDHMs) by using the quaternion algebra. Then the Rotation, Scaling and Translation (RST) invariants (CDHMIs and QDHMIs) are derived for

more description of images that have undergone affine transformations. In this paper we first illustrate the discrimination power of these moments by evaluating their efficiency in image reconstruction application. Then we propose a new approach for human face recognition based on these moment invariants (CDHMIs and QDHMIs) as descriptors and the Support Vector Machine (SVM) as supervised learning models that analyze data and recognize patterns. Experimental results, obtained using two public datasets, show that the proposed approach is more efficient when the disc-harmonic moments are used instead of other existing descriptors.

Keywords: disc-harmonic moments, moment invariants, quaternion algebra, image reconstruction, pattern recognition

Improving pedestrian detection using Support Vector Regression

Mounir ERRAMI, Mohammed RZIZA

Pedestrian detection has been always a chalenging problem in computer vision. Numerous approaches based on features extraction and classification have been proposed over the years. In this paper we present a novel pedestrian detection approach based on supervised classification. We propose here the use of basic statistical operators to adapt support vector regression (SVR) to binary classification. The classification chain adopted in this work is presented as follows : First, we use Haar wavelet decomposition and histograms of oriented gradients (HOG) for features extraction. For the classification task, we use the proposed SVR method and compare it with KNN and SVM classifiers. Experiments have been done on a public pedestrian data set. The obtained results prove the high performance of our proposed classification approach.

Keywords: Classification, features extraction, pedestrian detection, KNN, Support Vector Regression

CGiV2016_Abstract

39 | CGiV2016

A comparison on performance of data mining algorithms in classification of whewellite and weddellite urinary human calculi images color using different descriptors

AIT IDER Abdelouahed

The aim of this paper is to investigate the best correct rate of data mining supervised algorithms such as k-nearest neighbor (KNN), support vector machines (SVM), artificial neural networks (ANN), and PLS-DA. The significant information from the image is extracted using Zernike-, Gist-, and Centrist descriptor models in order to classify the two most frequent constituents of urinary stones using their colored images, calcium oxalate monohydrate and calcium oxalate dihydrate. The first class that contains images of calcium oxalate monohydrate (CaC2O4, H2O: whewellite (Wh)) crystal type. The second class contains the images of calcium oxalate dihydrate (CaC2O4, 2H2O weddellite (Wd)) crystal type. The results showed that the PLS-DA model coupled to Gist descriptor was superior to SVM-, kNN and KNN models in prediction using different descriptors. Using PLS-DA, SVM- KNN and ANN model the correct prediction rate reached 100%, 94.74%-, and 80.22% and 94.15% respectively.

Keywords: calcium oxalate; whewellite; weddellite; classification algorithms; Gist descriptor; Zernike moment

Session cgiv2014_3.3: CGiV - Application

Chair: Prof. Pun, Chi Man, University of Macau, Macau S.A.R.,

China

Modeling of 2D objects with weighted-Quadratic Trigonometric Spline

Muhammad Sarfraz, Shamaila Samreen, Malik Hussain

An imperative scheme is adopted to model 2D objects by constructing a weighted-spline using a quadratic trigonometric function with well controlled shape influences of parameters. The curve models, constructed

through the suggested scheme, own the best possible geometric properties such as convex hull, partition of unity, affine invariance and variation diminishing. The illustration of the method is supportive for various shape effects using interval tension property. The proposed C1 spline method is modest overall. Keywords: Bernstein Bézier form; trigonometric functions; weighted spline

Automatic Bifurcation Detection in Coronary X-Ray Angiographies

Asma Kerkeni, Asma BenAbdallah, Antoine Manzanera, Mohamed Hedi Bedoui

The detection of vascular bifurcation in X-ray images is important for several medical applications. They are used as landmarks for image registration, vessel segmentation and tracking. Although many bifurcation extraction methods have been proposed in recent years, very few work deals with coronary bifurcation in X-ray images. In this paper, we present a new bifurcation detector based on the multiscale Hessian analysis. It can be seen as a scale specific Histogram of Eigenvectors weighted by the vesselness measure. Pixels with three peaks in their immediate neighbourhood are considered as bifurcation candidates.

Based on this detector, a novel bifurcationness measure is proposed. The method is tested on real coronary artery angiographies and shows better results compared to other bifurcation detectors. Keywords: Hessian, Eigenvector, vesselness, Histogram, bifurcationness

CGiV2016_ NOTES

NOTES

F.T. Marchese, E. Banissi (Eds.)

Knowledge Visualization Currents

From Text to Art to Culture

▶ Presents the state of the art in visualization research and development

▶ Highlights research developing at key intersections with other disciplines and

its applicability to addressing complex real-world problems

▶ Discusses how visualization researchers are addressing complex issues of

representation in knowledge, art, and culture

Lying at the intersection of education, art, and cultural heritage, visualization is a powerful tool for representing and interpreting complex information. This unique text/reference reviews the evolution of the field of visualization, providing innovative examples of applied knowledge visualization from disciplines as varied as law, business management, the arts and humanities. With coverage of theoretical and practical aspects of visualization from ancient Sumerian tablets through to twenty-first century legal contracts, this work underscores the important role that the process of visualization plays in extracting, organizing, and crystallizing the concepts found in complex data. Topics and features:

• Contains contributions from an international selection of preeminent authorities • Presents a thorough introduction to the discipline of knowledge visualization, its current state of

affairs and possible future developments • Examines how tables have been used for information visualization in historical textual documents • Discusses the application of visualization techniques for knowledge transfer in business

relationships, and for the linguistic exploration and analysis of sensory descriptions • Investigates the use of visualization to understand orchestral music scores, the optical theory

behind Renaissance art, and to assist in the reconstruction of an historic church • Describes immersive 360 degree stereographic visualization, knowledge-embedded embodied

interaction, and a novel methodology for the analysis of architectural forms This interdisciplinary collection of the state of the art in knowledge visualization will be of considerable interest to researchers from a broad spectrum of backgrounds in both industry and academia.

http://www.graphicslink.co.uk/IV2016/

Swarms of ants and birds set off to decode a complex painting by Willem De

Kooning in their own swarmic way. The step-by-step behaviour of the swarms

is detailed in.

This video displays consecutive cycles of the behaviour of the swarms. A

cycle begins when swarms choose a line from the canvas on the right and

produce their interpretation of that line on the canvas on the left . Once they

have reproduced all the lines, they start a new cycle where they use the

outline of what they have just created to generate a new interpretation of it.

Over time, the swarms eat away the lines, simplifying the structure of the

original painting. This leads to a simpler line structure which ultimately

results in an empty canvas.

This artwork originated from an exchange with Professors William Latham,

Frédéric Fol Leymarie and Mark Bishop.

D I G I T A L A R T G A L L E R Y

o n l i n e e x h i b i t i o n July 2015 ~ June 2016

V I R T U A L G A L L E R Y V E N U E

www.g raph ic s l i nk .c o .uk /DART.h tm

© Mohammad Majid al-Rifaie ~ "Decoding De Kooning" -

©

G

abri

ele

Pet

ers

~D

ark

Day

s N

ew Y

ork

2"

"Inner Shrine": Inspired by the poem "Journey Home" composedby the noted Indian poet, Rabindranath Tagore, combined with my writing, "Inner Shrine" reveals the emotional and spiritual journey in searching of one's soul. To find the inward space of one's own- the home of the heart, one has to experience life to the fullest. It is the most intricate encounter that wanderers have to undertake, through which they discover the essence of their existence and the answer for who we

are. © Jing Ziou

Published by: Visualisation & Graphics Research Unit

London South Bank University

beni mellal city morocco ... · cgiv2016 _ doctoral research workshop cgiv2016 forum - ii 8th...

Documents