Evolving object detectors with a GPU accelerated vision system

Abstract

Using GPU processing, it is now possible to develop an evolutionary vision system working at interactive frame rates. Our system uses motion as an important cue to evolve detectors which are able to detect an object when this cue is not available. Object detectors consist of a series of high level operators which are applied to the input image. A matrix of low level point operators are used to recombine the output of the high level operators. With this contribution, we investigate, which image processing operators are most useful for object detection. It was found that the set of image processing operators could be considerably reduced without reducing recognition performance. Reducing the set of operators lead to an increase in speedup compared to a standard CPU implementation. © 2010 Springer-Verlag Berlin Heidelberg.