This paper proposes a new neural algorithm to perform the segmentation of an observed scene into regions corresponding to different moving objects byanalyzing a time-varying images sequence. The method consists of a c...
Recently, a qualitative approach was proposed for 3-D shape recovery based on a hybrid object representation[1]. In this approach, aspect recovery is the most important stage which binds regions in the image into mean...
This paPer addresses the issue of building a case-based preliminary design system by using Hopfield networks. one limitation of Hopfield networks is that it cannot be tralned, i.e. the weights between two neurons must...