Reaction-diffusion cellular nonlinear networks for feature enhancement in dermatoscopic images

Reaction-diffusion cellular neural networks have been studied because of their properties regarding processing time and capability of integration. The possibility of using them in image preprocessing and/or processing and the advantages and disadvantages of doing that are analyzed. The main objective is to find suitable sets of parameters (genes) that can enhance some useful properties of a dermatoscopic image. These properties should later help in processing and classifying images into healthy or suspect.