Real-time morphology processing using highly parallel 2D cellular automata CAM/sup 2/

This paper proposes a real-time morphology processing method based on a highly-parallel 2D cellular automata called CAM/sup 2/ (cellular automata on content addressable memory) that can attain pixel-order parallelism (several hundred thousands) on a single PC board. Evaluation results show that CAM/sup 2/ can perform various types of morphology processing within video rates, indicating it could be a promising platform for processing morphology.