A Digital Vision Chip for Early Feature Extraction with Rotated Template-Matching CA

We discuss a cellular-automata (CA) LSI core that extracts early features of objects in images, such as sizes and skeletons. A CMOS-image sensor with a CA core enables high-speed image processing. We propose an efficient CA algorithm based on rotated template matching. Each cell circuit in the proposed CA is implemented by a digital circuit, and transistors in each cell circuit number 198 in full customized design. The CA LSI consists of a large number of cell circuits operating in parallel to ensure fast, efficient object extraction as the number of cells increases. With a 0.25μm CMOS process, the total area of each cell circuit is 30×30μm². Simulation results indicated that image processing with 320×240 cells operates at up to 25MHz.