Retinomorphic vision systems

Publisher Website
Google Scholar
Abstract
The new generation of silicon retinae has two defining characteristics. First, these synthetic retinae are morphologically equivalent to their biological counterparts---at an appropriate level of abstraction. Second, they accomplish _all_ four major operations performed by biological retinae using neurobiological principles: (1) continuous sensing for detection, (2) local automatic gain control for amplification, (3) spatiotemporal bandpass filtering for preprocessing, and (4) adaptive sampling for quantization. I introduce the term _retinomorphic_ to refer to this subclass of the neuromorphic electronic systems [Mead90]. I compare and contrast their design principles with the standard practice in imager design. I argue that neurobiological principles are best suited to perceptive systems [Vittoz94] that go beyond reproducing the dynamic scene, like a conventional video camera does, to extracting salient information in real time [Andreou95]. I shall present results from a fully operational retinomorphic vision system and discuss the trade-offs involved in its design.

This publication has 31 references indexed in Scilit: