1 / 27

Transfer Color to Grayscale Image

Transfer Color to Grayscale Image. Review:. +. =. Another Example. +. =. Current Literature. Most techniques – Involve heavy human interaction Pseucoloring Movie Industry. Pseudocoloring. c(x, y) = T(f(x, y)) Choice of colormap – determined by human decision.

maree
Download Presentation

Transfer Color to Grayscale Image

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Transfer Color to Grayscale Image

  2. Review: + =

  3. Another Example + =

  4. Current Literature • Most techniques – Involve heavy human interaction • Pseucoloring • Movie Industry

  5. Pseudocoloring c(x, y) = T(f(x, y)) Choice of colormap – determined by human decision

  6. Movie Industry -- Silberg [1998] • Polygonize the image • Color individual components like a color book – human interaction • Track polygons between frames and transfer colors

  7. Reinhard et al. [2001] • Transfer color between two color images • Inspire the algorithm in Welsh’s paper

  8. Main Idea R G Intensity or Luminance B

  9. l Color Space • Goal: minimize correlation between three coordinate axes of the color space • l: luminance channel • : yellow-blue channel • : red-green channel • l color space provides a decorrelated achromatic channel • Transfer chromatic  and  channels from color image to grayscale image without cross-channel artifacts

  10. Algorithm (Details) • Step 1:

  11. RGB l color space

  12. Luminance Remapping • linearly shift and scale the luminance histogram of the source image to fit the histogram of the target image

  13. Algorithm (cont.) • Step 2: Jittered Sampling: 200

  14. Neighborhood Statistics • Weighted average of luminance (50%) and standard deviation (50%) • Neighborhood size: 5x5 pixels

  15. General Algorithm (cont.) • Step 3: ,  channel

  16. l RGB color space Step 4:

  17. Algorithm (cont.) • They also suggest: Swatch + =

  18. Results, Analysis & Discussion + =

  19. Sample Size + = 64x64 30x20 20x10

  20. Neighborhood Size + = 5x5 10x10

  21. Luminance Weight 50% + = 80% 30%

  22. Image Size 64x64 128x128 256x256

  23. Reasons for Problematic Area • Texture correspondence • - Standard Deviation This leads to the extension to SWATCH

  24. Texture correspondence + = + =

  25. Rounding Error

  26. Running Time • 15 seconds to 4 minutes on a Pentium III 900 MHz CPU using optimized MATLAB code • Ours…a few seconds on a Pentium III 1 GHz CPU using optimized VC++ code but…We are using different sized images, so… No comparison!

  27. References • GONZALEZ, R. C. AND WINTZ, P., 1987. Digital Image Processing, Addison-Wesley Publishing, Reading MA • HERTZMANN, A., JACOB, C., OLIVER, N., CURLESS, B., SALESIN., D., 2001. Image Analogies, Proceedings of ACM SIGGRAPH 2002, 341-346 • PITAS, I., Digital Image Processing Algorithms, Prentice Hall, 1993 • REINHARD, E., ASHIKHMIN, M., GOOCH B. AND SHIRLEY, P., 2001, Color Transfer between Images, IEEE Computer Graphics and Applications, September/October 2001, 34-40 • SILBERG, J., 1998. Cinesite Press Article, tttp://www.cinesite.com/core/press/articles/1998/10_00 _98-team.html • WELSH, T., ASHIKHMIN, M. AND MUELLER, K., 2002, Transferring Color to Greyscale Images, SIGGRAPH 2002, 277 - 280

More Related