An expert module to recognize injured rat footprint images

Abstract

Many medical research centers need to analyse rat footprint images in order to evaluate some nerve lesions. The authors present a system, called RER, which is able to analyse rat footprint images and detect fingers automatically. This system is composed of a low level that transforms images into structured information and a high level to recognize fingers in the footprint frames. Techniques employed by the high level to select fingers in a set of marks including noise and fingerprints are presented.<>