| 摘要 |
<p>An image encoder comprises an atom extracting unit (109) for acquiring position information representing the position of an atom of prediction residual data in each macro block and a variable-length encoding unit (113) for determining an occurrence frequency distribution of position information in each macro block to be encoded according to the number of atoms of the prediction residual data present in the macro block and arithmetically encoding the position information according to the thus determined occurrence frequency distribution. Thus, the efficiency of the entropy encoding of the atom parameter of the matching pursuit encoding is improved.</p> |
