Title :
Minimization of color image entropy through colorimetric transformation
Author :
Charrier, Christophe ; Cherifi, Hocine
Author_Institution :
Inst. d´´Ingenierie de la Vision, Jean Monnet Univ., Saint-Etienne, France
Abstract :
Image compression plays an important role in making digital images useful. Good compression systems must provide a trade-off between quality of representation for human beings and bit rate. We examine how entropy depends on choice of color space. We study entropy in both theoretical distributions of color image RGB values and distributions drawn from natural images. We determine which of several color spaces has minimal entropy, including several that purport to model human color processing. Our analytical results are supported by results found with a variety of images. Both analytical results and empirical results show that the UVW color space best minimizes entropy
Keywords :
Gaussian distribution; colorimetry; colour vision; image coding; image colour analysis; image representation; minimum entropy methods; visual perception; Gaussian distribution; UVW color space; bit rate; color image RGB values distribution; color image entropy minimization; color space; colorimetric transformation; compression systems; digital images; human color processing; image compression; image quality; image representation; minimal entropy; natural images; visual perception; Bit rate; Digital images; Entropy; Equations; Humans; Image analysis; Image coding; Image color analysis; Psychology; Visual perception;
Conference_Titel :
Digital Signal Processing Workshop Proceedings, 1996., IEEE
Conference_Location :
Loen
Print_ISBN :
0-7803-3629-1
DOI :
10.1109/DSPWS.1996.555449
