Fuzzy processing for active vision

Humans employ an active visual system to gather visual data from the surrounding environment. In addition to continually re-focusing the lens and adjusting the iris to control the exposure of light on to the retina, the eye constantly moves so that specific information can be focused on to the fovea – the part of the retina capable of defining fine detail. The target location when the eye moves is determined by the attention of the viewer. The attention of the viewer is drawn towards regions within the environment that are interesting or salient, i.e.̃they contain visual or semantic information. A mobile robot traversing within its environment has to identify and locate landmarks and other mobile entities as part of its navigational processes. The ability to efficiently analyze visual data captured from the environment and relate this to information from its internal map is very important. In this work a technique is described that uses semantic data to direct attention towards regions of interest within a simple environment and systematically search other areas of the environment. A fuzzy system lies at the heart of the technique that reasons with historical and environmental data in order to influence the search patterns. A Kohonen self-organizing map is used for object classification. Experimental data will be provided to illustrate the success and efficiency of the technique and conclusions will be drawn that discuss future work.