| 摘要 |
<p>PURPOSE:To efficiently reduce the redundancy of picture data by allocating a shorter code word in the sequence from a forecasting error with a high generation probability. CONSTITUTION:A picture data processor is composed of a picture data dividing processing part 10, a reference concentration value detecting part 20, a concentration change calculating part 30, a forecasting concentration value calculating part 50, a storing part 60 and an encoding processing part 70. Here, the original picture data are divided for a micro block composed of plural adjoining picture elements, the concentration value of a color component except a reference color is forecasted concerning respective divided blocks and the error (forecasting error) between the forecasted concentration value and the actual concentration value of the block is obtained. Thereafter, the code word is suitably allocated to the reference concentration value of respective blocks and respective concentration values in the block of a reference color, and the shorter code word is allocated to the forecasting error in the sequence of the high generation probability. Thus, the redundancy of the picture data can be efficiently reduced.</p> |
