摘要 |
A method of decompressing a document image compressed with JPEG ADCT and a subsequent statistical encoding method, includes an iterative processing step, which filters (320) the image to correct for compression/decompression artifacts, and subsequently checks (322,324) the ADCT coefficients of the filtered image to determine whether the filtered image is a possible one, given the received original ADCT coefficients and Q-Table description. In such an arrangement, the iterative process step (312,314) can be terminated at a preselected number of iterations (304) based on the measured size (302) of the statistical encoding prior to removing (306) the statistical encoding. <IMAGE> |