The use of entropy for colour edge detection
A. Fotinos, George Economou, Evangelos Zigouris, S. Fotopoulos · 2003
The edge detection of images is a widely discussed topic and several methods have been presented. In this work the vector nature of the color image is retained assuming the RGB space as the standard processing space. The proposed statistical approach is based on entropy and relative entropy operator using the Parzen kernels to estimate the local density distribution. Experimental results indicate that these methods, if not better, are comparable to other standard colour edge detection techniques.