Presentation is loading. Please wait.

Presentation is loading. Please wait.

Color Imaging 2004 1 Analysis of Spatio-chromatic Decorrelation for Colour Image Reconstruction Mark S. Drew and Steven Bergner

Similar presentations


Presentation on theme: "Color Imaging 2004 1 Analysis of Spatio-chromatic Decorrelation for Colour Image Reconstruction Mark S. Drew and Steven Bergner"— Presentation transcript:

1 Color Imaging 2004 1 Analysis of Spatio-chromatic Decorrelation for Colour Image Reconstruction Mark S. Drew and Steven Bergner {mark/sbergner}@cs.sfu.ca School of Computing Science, Simon Fraser University, Canada

2 Color Imaging 2004 2 2/27 - Use of PCA vs. ICA — what’s the difference? - How do you do ICA? - What does this have to do with images? - The objective: best characterize image blocks using ICA on color image block data == spatio (blocks are 16x16, say)- chromatic (x3); assign bits in bit allocation according to the importance of each ICA coefficient  data compression. I. Overview

3 Color Imaging 2004 3 3/27 Best characterize image  colour and spatial information. Colour: we think of using PCA (Principal Component Anaysis): discover main colour axes. Is this best, given our objective? Spatial: use spatial Fourier filters? Gabor wavelets? Etc. Here, we’ll use ICA (Independent Component Anaysis) to derive best colour and spatial decomposition at once, for decorrelation, compression, and reconstruction.

4 Color Imaging 2004 4 4/27 II. ICA  What is it? ICA is a form of “Blind Source Separation”  To explain, consider audio signals (in an Imaging conference!). Consider 2 speakers, and 2 microphones: s1s1 s2s2 -sources x1x1 x2x2 -data

5 Color Imaging 2004 5 5/27 Can we disentangle s 1, s 2 from measured data x 1, x 2 ? == The “cocktail party problem”. An example:

6 Color Imaging 2004 6 6/27 ICA: Order and sign not determined.

7 Color Imaging 2004 7 7/27 What about PCA? Writing the signals in terms of reduced set of sources s 1, s 2, s 3,..., for higher-dimensional data, we can do a better job in compression. 

8 Color Imaging 2004 8 8/27 III. ICA  How to do it? Model: ( x was 2xN in the audio example.) mixing matrix separating matrix

9 Color Imaging 2004 9 9/27 Driving idea for finding sources: s 1, s 2 are statistically independent == information about one gives no knowledge re. the other. Not just uncorrelated: covariance = 0 ==PCA

10 Color Imaging 2004 10 10/27 If independent as well, the pdf is separable: joint pdf marginal pdf’s which implies for any functions, !  useful for solving.

11 Color Imaging 2004 11 11/27 So, to do ICA, start with uncorrelated signals (using PCA) == simplifies. Main tool: Non-Gaussian is independent. Central Limit Theorem: the sum of two independents is more like a Gaussian than is either one. So  we have sums. To get s, make a linear combination of x ’s that is as non-Gaussian as possible.

12 Color Imaging 2004 12 12/27 One way: (…many others) A Gaussian has zero kurtosis. For zero mean y, Rescale y to variance=1:  just use We seek a signal that maximizes kurtosis.

13 Color Imaging 2004 13 13/27 Algorithm  “whiten” the data: zero mean, + linear transform to make uncorrelated, variance=1. First, PCA: orthogonal U with In the new coordinate system, Why?  Now with orthogonal  simpler to search for.

14 Color Imaging 2004 14 14/27 Algorithm -whiten x -we seek a column w of orthogonal W, with, that maximizes kurtosis: Euler eqn.: Code 1. Initialize w randomly, with 2. 3. 4. stop when

15 Color Imaging 2004 15 15/27 Matlab

16 Color Imaging 2004 16 16/27 IV. ICA for Images Previous work: Greyscale and colour imagery using PCA and ICA. For colour images, x could be 3-vector pixels. But get spatial as well if use n  n tiles (nice illustration in Süsstrunk et al., CGIV’04 [using PCA on raw CFA data]) We show here that compression is better using ICA+colour+spatial info.

17 Color Imaging 2004 17 17/27 16 x 16 greyscale tiles ICA finds “sparse” features: ICA (16 2 x1 greyscale data) localization in space

18 Color Imaging 2004 18 18/27 PCA vs. ICA (3x1 data) (no spatial information) With colour:

19 Color Imaging 2004 19 19/27 PCA (4x4 x3) DCT (4x4 x3) -less axis-aligned -ordering by variance-accounted- for is different: pure colour axes appear first -pure colour axes appear later, after luminance frequencies -separates colour from luminance PCA vs. DCT (4x4 x3 data) -Colour: luminance, blue-yellow, red-green

20 Color Imaging 2004 20 20/27 ICA (4x4 x3) PCA (4x4 x3) again PCA vs. ICA -colour less separate from spatial information -combined localization in space and frequency -patterns not rectangular  more like Gabor functions (Gaussian-modulated sine functions) -localization in frequency

21 Color Imaging 2004 21 21/27 ICA (4x4) ICA (5x5) ICA (8x8) ICA (16x16)

22 Color Imaging 2004 22 22/27 SNRSNR Colour vs. Greyscale:  Compression  performance (Generic basis) Colour Greyscale - Higher reconstruction quality (SNR) for larger patches - Colour has better quality than grey, at equal compression Better quality 

23 Color Imaging 2004 23 23/27 ICA vs. PCA (Specific basis: image = ) - ICA much better than PCA: higher compression for same SNR - ICA  increased quality with larger patches, for equal compression ICA PCA Better quality 

24 Color Imaging 2004 24 24/27 ICA vs. PCA A. ICA does better separating axes such that they influence each other least  better entropy coding B.Colour aids in compression C.Large patch sizes and low rate encoding  At equal compression, SNR (quality) better for ICA

25 Color Imaging 2004 25 25/27 ICA vs. PCA: Image reconstruction (compression ratio: 1:12) ICA PSNR= 35.55 DCT: PSNR= 31.97

26 Color Imaging 2004 26 26/27 Another image ICA PSNR= 39.69 DCT: PSNR= 31.40  7:1 Orig ICA DCT --blocking

27 Color Imaging 2004 27 27/27 The Future: Video Bases [submitted] ICA (6x6x6) PCA (6x6x6)


Download ppt "Color Imaging 2004 1 Analysis of Spatio-chromatic Decorrelation for Colour Image Reconstruction Mark S. Drew and Steven Bergner"

Similar presentations


Ads by Google