Hypercomplex Auto-And-Cross-Correlation of Color Images

10 years 2 months ago
Hypercomplex Auto-And-Cross-Correlation of Color Images
Autocorrelation and cross-correlation have been defined and utilized in signal and image processing for many years, but not for color or vector images. In this poster we present for the first time a definition of correlation applicable to color images, based on quaternions or hypercomplex numbers. We have devised a visualization of the result using the polar form of a quaternion in which color denotes quaternion eigenaxis and phase, and a grayscale image represents the modulus. 3. Color image representation A color image in rgb color space may be represented using hypercomplex numbers by encoding the red, green and blue components of the rgb value as a pure quaternion: f(x, y) = ir(x, y) + jg(x, y) + kb(x, y) where r(x, y) is the red component of the color image and similarly for the green and blue components. The reason for choosing this representation is that the rgb values represent a 3-space vector (a point in rgb space), as does the pure quaternion. 6. Results ? autocorrelation o...
Stephen J. Sangwine, Todd A. Ell
Added 25 Oct 2009
Updated 26 Oct 2009
Type Conference
Year 1999
Where ICIP
Authors Stephen J. Sangwine, Todd A. Ell
Comments (0)