ABSTRACT Numerous successful methodologies have been developed for grey scale image processing and analysis. However, theirapplications to color images are plagued because of the multi-dimensionality and inter-correlation of color images. In thispaper, we examine the decorrelation of color images using the principal component analysis. Both global and local analysesare studied. To more efficiently decorrelate the images based on local features, a total color difference measurement is utilizedto find the regions in which the decorrelation takes place. We call this dynamic decorrelation. Experiments are enclosed andcomparisons of dynamic decorrelation with both global and local decorrelation are conducted. 1. Introduction In the past, the area of computer vision has predominantly dealt with the processing of either binary or grey scale images,with no or little emphasis placed on color [1,2,14]. Consequently most existing vision systems are "color blind". For someparticular objects in an unstructured environment, such as those with the same luminosity from different objects or background,