Gpu Implementation Of A Bio-Inspired Vision Model

This paper describes the implementation of a bio-inspired visual processing model which aims to remark the relevant information from a scene to generate different kinds of stimuli such as visual, haptic or audio stimuli, useful in aid systems for visually impaired persons. Desirable features for such system are real-time operation, minimum latency and portability. To meet all these requirements we have implemented the whole model in the GPU NVIDIA ION2 which is integrated in a light-weight netbook and performs at a maximum rate of 35 frames per second.