Color fractal images with correlated RGB color components

The color fractal images with correlated RGB color components were generated using the midpoint displacement alogrithm, using vectorial increments in the RGB color space, according to a multivariate Gaussian distribution specified by the variance-covariance matrix. This data set contains two sets of 25 color fractal images with two color components, of varying complexity expressed as the color fractal dimension, as a function of (i) the Hurst coefficient that was varied from 0.1 to 0.9 in steps of 0.2 and (ii) the correlation coefficient between the red and green color channels. It also contains a set of 9 color fractal images with three color components, for an arbitrary configuration of the variance-covariance matrix. Each fractal object was either independently rendered as a color image (the not normalized set) or applying a global normalization by considering the extreme values of the most complext color fractal image (the normalized set). The data set is intented to be used as a reference data set for color texture complexity analysis when considering fractal dimension estimation.