image processing & pattern recognition automated solar cavity detection 1 athena johnson

32
IMAGE PROCESSING & PATTERN RECOGNITION AUTOMATED SOLAR CAVITY DETECTION 1 Athena Johnson

Upload: charla-neal

Post on 13-Dec-2015

218 views

Category:

Documents


0 download

TRANSCRIPT

  • Slide 1

IMAGE PROCESSING & PATTERN RECOGNITION AUTOMATED SOLAR CAVITY DETECTION 1 Athena Johnson Slide 2 OUTLINE Introduction Background Problem Statement Proposed Solution Experiments Conclusions Future Work 2 Slide 3 INTRODUCTION 3 Slide 4 BACKGROUND Solar Dynamics Observatory (SDO) Extreme Ultraviolet Variability Experiment (EVE) Helioseismic and Magnetic Imager (HMI) Atmospheric Imaging Assembly (AIA) 1.5 Terabytes (TB) of data per day 4 Slide 5 Atmospheric Imaging Assembly (AIA) Images the Corona of the Sun Study of solar storms How they are created? How they propagate upward? How they emerge from the Sun? How magnetic fields heat the corona? 5 Slide 6 SOLAR CAVITIES Currently an increase in implementations focused on Solar Cavities Off limb structures Darker elliptical structure, encompassed by lighter regions Hypothesized to be precursors to solar events Aid in establishing a predictive solar weather system 6 Slide 7 SOLAR CAVITIES Labrosse, Dalla and Marshall (2010) Radial intensity profiles Support Vector Machine (SVM) Region growing Calculation of metrics Running difference on subsequent images 7 Slide 8 SOLAR CAVITIES Durak and Nasraoui (2010) Exraction of principal contours Calculations on contours Adaboost 8 Slide 9 PROBLEM STATEMENT Computation times Detections based on metrics Weak events missed Multiple detections Multiple events missed Low hit rates 9 Slide 10 HAAR CLASSIFIER Method that Paul Viola and Michael Jones published in 2001 Four key concepts Haar-like features Integral Image Adaboosting Cascade of Classifiers 10 Slide 11 HAAR-LIKE FEATURES Aids in satisfying real time requirements Rectangular regions Reduces Computation 11 Slide 12 INTEGRAL IMAGES Rapid computation of Haar-like features 12 Slide 13 INTEGRAL IMAGES Original Image 8+6+2+5+6+3 = 30 Integral Image 50-17-5+2 = 30 13 Slide 14 ADABOOSTING Aids in increasing the accuracy and speed Begins with uniform weights over training examples Obtain a weak classifier Update weights 14 Weak Classifier h1(x) Slide 15 ADABOOSTING 15 Weak Classifier h2(x) Weak Classifier h3(x) Slide 16 ADABOOSTING Weak classifiers combined to form the strong classifier 16 Slide 17 CASCADE OF CLASSIFIERS Increases the speed of detections All Haar-like features from all stages combined into a final Classifier Model Cascade of boosted classifiers with Haar-like features 17 Slide 18 CASCADE OF CLASSIFIERS A series of classifiers are applied to every subwindow of image A positive result from the first classifier, triggers evaluation from the second classifier and so on 18 Slide 19 INITIAL SOLUTION 19 Slide 20 RESULTS Manually selected Training Image Sets Positive Samples = 100 Negative Samples = 400 79.6% Correct detection rate was achieved 20 Slide 21 RESULTS Missed detections in specific quadrants Detections on the Suns disk Overlapping detections 21 Slide 22 PROPOSED SOLUTION 22 Slide 23 MINIMIZED TRAINING SETS 10 Positive Images10 Negative Images 23 Slide 24 MARK REGIONS OF INTEREST AND ROTATE Deriving images from selected images Rotation applied to both training sets 24 Slide 25 TRANSFORM REGIONS OF INTEREST Transformations on cavities 25 Slide 26 PREPROCESSING Edge Detection Hough Lines Calculate the radius 26 Slide 27 RESULTS Derived Training Image Sets Initial image in sets = 10 Positive Samples = 3600 Negative Samples = 3600 96% Correct detection rate was achieved 27 Slide 28 FINAL IMAGE WITH DETECTIONS 28 Slide 29 CONCLUSION Less manual work Short training times < 22 hours Wider range of detections Weak and strong cavities Fast run times < 1 second per image Higher hit rates 29 Slide 30 FUTURE WORK Technique Improvement Reduction of False Positives Contour Detections Template Matching Customized Haar-like features 30 Slide 31 FUTURE WORK Find optimal number of training sets Extract Metrics User Interface 31 Slide 32 32 QUESTIONS?