1 algorithms compared bicubic interpolation mitra's directional filter fuzzy logic filter...
TRANSCRIPT
![Page 1: 1 Algorithms compared Bicubic Interpolation Mitra's Directional Filter Fuzzy Logic Filter Vector Quantization VISTA](https://reader036.vdocuments.us/reader036/viewer/2022062511/551479e1550346b0158b548b/html5/thumbnails/1.jpg)
1
Algorithms compared
• Bicubic Interpolation• Mitra's Directional Filter• Fuzzy Logic Filter•Vector Quantization• VISTA
![Page 2: 1 Algorithms compared Bicubic Interpolation Mitra's Directional Filter Fuzzy Logic Filter Vector Quantization VISTA](https://reader036.vdocuments.us/reader036/viewer/2022062511/551479e1550346b0158b548b/html5/thumbnails/2.jpg)
2
Bicubic spline Altamira VISTA
![Page 3: 1 Algorithms compared Bicubic Interpolation Mitra's Directional Filter Fuzzy Logic Filter Vector Quantization VISTA](https://reader036.vdocuments.us/reader036/viewer/2022062511/551479e1550346b0158b548b/html5/thumbnails/3.jpg)
3
Bicubic spline Altamira VISTA
![Page 4: 1 Algorithms compared Bicubic Interpolation Mitra's Directional Filter Fuzzy Logic Filter Vector Quantization VISTA](https://reader036.vdocuments.us/reader036/viewer/2022062511/551479e1550346b0158b548b/html5/thumbnails/4.jpg)
4
User preference test results
“The observer data indicates that six of the observers rankedFreeman’s algorithm as the most preferred of the five testedalgorithms. However the other two observers rank Freeman’s algorithmas the least preferred of all the algorithms….
Freeman’s algorithm produces prints which are by far the sharpestout of the five algorithms. However, this sharpness comes at a priceof artifacts (spurious detail that is not present in the originalscene). Apparently the two observers who did not prefer Freeman’salgorithm had strong objections to the artifacts. The other observersapparently placed high priority on the high level of sharpness in theimages created by Freeman’s algorithm.”
![Page 5: 1 Algorithms compared Bicubic Interpolation Mitra's Directional Filter Fuzzy Logic Filter Vector Quantization VISTA](https://reader036.vdocuments.us/reader036/viewer/2022062511/551479e1550346b0158b548b/html5/thumbnails/5.jpg)
5
![Page 6: 1 Algorithms compared Bicubic Interpolation Mitra's Directional Filter Fuzzy Logic Filter Vector Quantization VISTA](https://reader036.vdocuments.us/reader036/viewer/2022062511/551479e1550346b0158b548b/html5/thumbnails/6.jpg)
6
![Page 7: 1 Algorithms compared Bicubic Interpolation Mitra's Directional Filter Fuzzy Logic Filter Vector Quantization VISTA](https://reader036.vdocuments.us/reader036/viewer/2022062511/551479e1550346b0158b548b/html5/thumbnails/7.jpg)
7
Training images
![Page 8: 1 Algorithms compared Bicubic Interpolation Mitra's Directional Filter Fuzzy Logic Filter Vector Quantization VISTA](https://reader036.vdocuments.us/reader036/viewer/2022062511/551479e1550346b0158b548b/html5/thumbnails/8.jpg)
8
Training image
![Page 9: 1 Algorithms compared Bicubic Interpolation Mitra's Directional Filter Fuzzy Logic Filter Vector Quantization VISTA](https://reader036.vdocuments.us/reader036/viewer/2022062511/551479e1550346b0158b548b/html5/thumbnails/9.jpg)
9
Processed image
![Page 10: 1 Algorithms compared Bicubic Interpolation Mitra's Directional Filter Fuzzy Logic Filter Vector Quantization VISTA](https://reader036.vdocuments.us/reader036/viewer/2022062511/551479e1550346b0158b548b/html5/thumbnails/10.jpg)
10
Inference in Markov Random Fields
Gibbs sampling, simulated annealingIterated conditional modes (ICM)Belief propagation
Application examples:super-resolutionmotion analysisshading/reflectance separation
Graph cutsVariational methods
![Page 11: 1 Algorithms compared Bicubic Interpolation Mitra's Directional Filter Fuzzy Logic Filter Vector Quantization VISTA](https://reader036.vdocuments.us/reader036/viewer/2022062511/551479e1550346b0158b548b/html5/thumbnails/11.jpg)
11
Motion applicationimage patches
image
scene
scene patches
![Page 12: 1 Algorithms compared Bicubic Interpolation Mitra's Directional Filter Fuzzy Logic Filter Vector Quantization VISTA](https://reader036.vdocuments.us/reader036/viewer/2022062511/551479e1550346b0158b548b/html5/thumbnails/12.jpg)
12
What behavior should we see in a motion algorithm?
• Aperture problem
• Resolution through propagation of information
• Figure/ground discrimination
![Page 13: 1 Algorithms compared Bicubic Interpolation Mitra's Directional Filter Fuzzy Logic Filter Vector Quantization VISTA](https://reader036.vdocuments.us/reader036/viewer/2022062511/551479e1550346b0158b548b/html5/thumbnails/13.jpg)
13
The aperture problem
![Page 14: 1 Algorithms compared Bicubic Interpolation Mitra's Directional Filter Fuzzy Logic Filter Vector Quantization VISTA](https://reader036.vdocuments.us/reader036/viewer/2022062511/551479e1550346b0158b548b/html5/thumbnails/14.jpg)
14
The aperture problem
![Page 15: 1 Algorithms compared Bicubic Interpolation Mitra's Directional Filter Fuzzy Logic Filter Vector Quantization VISTA](https://reader036.vdocuments.us/reader036/viewer/2022062511/551479e1550346b0158b548b/html5/thumbnails/15.jpg)
15
Motion analysis: related work
• Markov network– Luettgen, Karl, Willsky and collaborators.
• Neural network or learning-based– Nowlan & T. J. Senjowski; Sereno.
• Optical flow analysis– Weiss & Adelson; Darrell & Pentland; Ju,
Black & Jepson; Simoncelli; Grzywacz & Yuille; Hildreth; Horn & Schunk; etc.
![Page 16: 1 Algorithms compared Bicubic Interpolation Mitra's Directional Filter Fuzzy Logic Filter Vector Quantization VISTA](https://reader036.vdocuments.us/reader036/viewer/2022062511/551479e1550346b0158b548b/html5/thumbnails/16.jpg)
16
Motion estimation results (maxima of scene probability distributions displayed)
Initial guesses only show motion at edges.
Iterations 0 and 1
Inference:
Image data
![Page 17: 1 Algorithms compared Bicubic Interpolation Mitra's Directional Filter Fuzzy Logic Filter Vector Quantization VISTA](https://reader036.vdocuments.us/reader036/viewer/2022062511/551479e1550346b0158b548b/html5/thumbnails/17.jpg)
17
Motion estimation results
Figure/ground still unresolved here.
(maxima of scene probability distributions displayed)
Iterations 2 and 3
![Page 18: 1 Algorithms compared Bicubic Interpolation Mitra's Directional Filter Fuzzy Logic Filter Vector Quantization VISTA](https://reader036.vdocuments.us/reader036/viewer/2022062511/551479e1550346b0158b548b/html5/thumbnails/18.jpg)
18
Motion estimation results
Final result compares well with vector quantized true (uniform) velocities.
(maxima of scene probability distributions displayed)
Iterations 4 and 5
![Page 19: 1 Algorithms compared Bicubic Interpolation Mitra's Directional Filter Fuzzy Logic Filter Vector Quantization VISTA](https://reader036.vdocuments.us/reader036/viewer/2022062511/551479e1550346b0158b548b/html5/thumbnails/19.jpg)
19
Inference in Markov Random Fields
Gibbs sampling, simulated annealingIterated conditional modes (ICM)Belief propagation
Application examples:super-resolutionmotion analysisshading/reflectance separation
Graph cutsVariational methods
![Page 20: 1 Algorithms compared Bicubic Interpolation Mitra's Directional Filter Fuzzy Logic Filter Vector Quantization VISTA](https://reader036.vdocuments.us/reader036/viewer/2022062511/551479e1550346b0158b548b/html5/thumbnails/20.jpg)
Forming an Image
Surface (Height Map)
Illuminate the surface to get:
The shading image is the interaction of the shapeof the surface and the illumination
Shading Image
![Page 21: 1 Algorithms compared Bicubic Interpolation Mitra's Directional Filter Fuzzy Logic Filter Vector Quantization VISTA](https://reader036.vdocuments.us/reader036/viewer/2022062511/551479e1550346b0158b548b/html5/thumbnails/21.jpg)
Painting the Surface
Scene
Add a reflectance pattern to the surface. Points inside the squares should reflect less light
Image
![Page 22: 1 Algorithms compared Bicubic Interpolation Mitra's Directional Filter Fuzzy Logic Filter Vector Quantization VISTA](https://reader036.vdocuments.us/reader036/viewer/2022062511/551479e1550346b0158b548b/html5/thumbnails/22.jpg)
Problem
What if we want to do shape from shading?
Need some mechanism of informing theshape from shading algorithm about thereflectance pattern.
Image Height Map
![Page 23: 1 Algorithms compared Bicubic Interpolation Mitra's Directional Filter Fuzzy Logic Filter Vector Quantization VISTA](https://reader036.vdocuments.us/reader036/viewer/2022062511/551479e1550346b0158b548b/html5/thumbnails/23.jpg)
Represent Scene As
These types of images are known as intrinsic images (Barrow and Tenenbaum)
=
Shading ImageImage Reflectance Image
![Page 24: 1 Algorithms compared Bicubic Interpolation Mitra's Directional Filter Fuzzy Logic Filter Vector Quantization VISTA](https://reader036.vdocuments.us/reader036/viewer/2022062511/551479e1550346b0158b548b/html5/thumbnails/24.jpg)
Not Just for Shape from Shading
• Ability to reason about shading and reflectance independently is necessary for most image understanding tasks– Material recognition– Image segmentation
• Intrinsic images are a convenient representation of the scene– Less complex than fully reconstructing the scene
![Page 25: 1 Algorithms compared Bicubic Interpolation Mitra's Directional Filter Fuzzy Logic Filter Vector Quantization VISTA](https://reader036.vdocuments.us/reader036/viewer/2022062511/551479e1550346b0158b548b/html5/thumbnails/25.jpg)
Goal
Image Shading Image Reflectance Image