Method and apparatus for processing both still and moving visual pattern images

Abstract

An improved method for coding and decoding still or moving visual pattern images by partitioning images into blocks or cubes, respectively, and coding each image separately according to visually significant responses of the human eye. Coding is achieved by calculating and subtracting a mean intensity value from digital numbers within each block or cube and detecting visually perceivable edge locations within the resultant residual sub-image. If a visually perceivable edge is contained within the block or cube, gradient magnitude and orientation at opposing sides of the edge within each edge block or cube are calculated and appropriately coded. If no perceivable edge is contained within the block or cube, the sub-image is coded as a uniform intensity block. Decoding requires receiving coded mean intensity value, gradient magnitude and pattern code, and then decoding a combination of these three indicia to be arranged in an orientation substantially similar to the original digital image or original sequence of digital images. Coding and decoding can be accomplished in a hierarchical pattern. Further, hierarchical processing can be programmably manipulated according to user-defined criteria.