GRADIENT PREDICTOR FOR IMAGE COMPRESSION

摘要

A method includes selecting, from an image, a target block of pixels to be compressed, selecting a neighbor column of the target block, selecting a neighbor row of the target block, generating two or more basis values based on at least one of a portion of pixels of the neighbor column and a portion of pixels of the neighbor row, calculating a gradient based on the two or more basis values, and assigning at least one predictor for the target block using the calculated gradient.