发明名称 COMPRESSION METHOD FOR MULTILEVEL IMAGE DATA
摘要 PURPOSE:To obtain a recovery image with high picture quality, by deciding the final quantizing width of the transform coefficient AC component of each block setting the generating frequency of the transform coefficient AC component quantized to zero in each block as a parameter. CONSTITUTION:Read out block data is cosine-transformed by a two-dimensional discrete consine transform device 3, and 256 transform coefficients of one block can be obtained. Next, 255 transform coefficient AC components except one DC component are quantized uniformly with fundamental quantizing width w0 by a quantizer 4, and a pair of a coefficient number that is a code with fixed length and a polarity number can be obtained at every transform coefficient. The pair of the coefficient number and the polarity number is stored in a block buffer memory 6 transiently, and also, the coefficient number is sent to a classifying device 5, and the classifying device 5 performs a classifying processing, and outputs a class number to ancoder 7. The coder 7 transforms a code(number) with fixed length of the transform coefficient AC component to a code with variable length. In such a way, it is possible to obtain the recovery image with high compressibility and high picture quality.
申请公布号 JPS63246977(A) 申请公布日期 1988.10.13
申请号 JP19870079775 申请日期 1987.04.02
申请人 KONICA CORP 发明人 YONEKAWA HISASHI;TAKEUCHI HIROSHI
分类号 A61B6/00;A61B6/03;G06T1/00;G06T9/00;H04N1/41;H04N19/126;H04N19/134;H04N19/136;H04N19/176;H04N19/189;H04N19/196;H04N19/423;H04N19/60;H04N19/625;H04N19/85;H04N19/91;H04N19/93 主分类号 A61B6/00
代理机构 代理人
主权项
地址