1. Texture Analysis and Synthesis

2. Texture Texture: pattern that “looks the same” at all locationsTexture: pattern that “looks the same” at all locations May be structured or randomMay be structured or random [Wei & Levoy]

3. Applications of Textures Texture analysisTexture analysis – Detemining statistical properties of textures – Segmentation – Recognition – Shape from texture Texture synthesisTexture synthesis

4. Oriented Filter Banks MultiresolutionOriented Filter Bank Original Image SteerablePyramid

5. Steerable Pyramid Texture Analysis Pass image through filter bankPass image through filter bank Compile histogram of intensities output by each filterCompile histogram of intensities output by each filter To synthesize new texture:To synthesize new texture: – Start with random noise image – Adjust histograms to match original image – Re-synthesize image from filter outputs

6. Texture Analysis / Synthesis OriginalTexture SynthesizedTexture Heeger and Bergen

7. Textons Elements (“textons”) either identical or come from some statistical distributionElements (“textons”) either identical or come from some statistical distribution Can analyze in natural imagesCan analyze in natural images Olhausen & Field

8. Clustering Textons Output of bank of n filters can be thought of as vector in n -dimensional spaceOutput of bank of n filters can be thought of as vector in n -dimensional space Can cluster these vectors using k -means [Malik et al.]Can cluster these vectors using k -means [Malik et al.] Result: dictionary of most common texturesResult: dictionary of most common textures

9. Clustering Textons Image Clustered Textons Texton to Pixel Mapping [Malik et al.]

10. Using Texture in Segmentation Compute histogram of how many times each of the k clusters occurs in a neighborhoodCompute histogram of how many times each of the k clusters occurs in a neighborhood Define similarity of histograms h i and h j using  2Define similarity of histograms h i and h j using  2 Different histograms  separate regionsDifferent histograms  separate regions

11. Texture Segmentation [Malik et al.]

12. Markov Random Fields Different way of thinking about texturesDifferent way of thinking about textures Premise: probability distribution of a pixel depends on values of neighborsPremise: probability distribution of a pixel depends on values of neighbors Probability the same throughout imageProbability the same throughout image – Extension of Markov chains

13. Texture Synthesis Based on MRF For each pixel in destination:For each pixel in destination: – Take already-synthesized neighbors – Find closest match in original texture – Copy pixel to destination Efros & Leung 1999, speedup by Wei & Levoy 2000Efros & Leung 1999, speedup by Wei & Levoy 2000 Extension to copying whole blocks by Efros & Freeman 2001Extension to copying whole blocks by Efros & Freeman 2001 [Wei & Levoy]