Image Recognition Processor based on Morphological Associative Memories

In this paper, the description of an image recognition processor based on morphological associative memories (MAM) is presented. The combination of the MAM features with its implementation in a programmable logic device grants to the system high speeds of processing, high immunity to the present noise in the images and great capacity of storage; these features make of our proposal a robust system and ideal to work in applications with autonomous real time systems. The described processor, with aid of a binary process of the image, realizing the filter functions, and the use of morphological heteroassociative memories min demonstrated a high performance in the recognition of images corrupted with diverse types of noises.