A new SIMD computer vision architecture with image algebra programming environment

New programmable hardware architectures and much more efficient algorithm development environment tools are needed to meet the explosive growth and diverse application scenarios anticipated for computer vision technology over the next 5 years in both the military and commercial application arenas. Needed are programmable computer architectures that provide much higher cost effectiveness through scalability of all hardware resources (i.e., I/O bandwidth, throughput, memory) and through exploitation of the unique requirements and characteristics of image processing algorithms and applications. In addition, a new class of software development tools is needed that can streamline and improve the efficiency of the algorithm development process, as well as provide a concise, platform-independent, very-high-level notation for computer vision algorithms that encourages iterative and collaborative developments. Progress in developing new tools and technology for military and commercial computer vision applications by the Parallel Algebraic Logic (PAL) consortium is described