Segmentation coding for object-based attentive selection systems

In this paper we present an analysis of a segmentation coding algorithm used in object-based visual attentive selection systems. This coding algorithm segments images into objects, and assigns a segmentation code to each object. The objects then compete for attention in the selection process. Thus, these systems use a biologically-inspired technique to balance the need for high-speed processing and the abundance of input image data. We have implemented these systems in 2.0 /spl mu/m and 1.2 /spl mu/m CMOS processes.