| 摘要 |
<p>PURPOSE:To automatically decrease compressed data smaller than a line allowable quantity by providing a quantization table selecting part and selecting a quantization table based on the sum of the bit number of a code block. CONSTITUTION:A two-dimensional orthogonally intersecting conversion part 1 applies image data two-dimensional orthogonal conversion. The 2-dimensional orthogonal coefficient obtained from the result is quantized by referring to a quantization threshold by a quantization part 9. The quantized quantization- coefficient is encorded by an encoding part 13. A standard quantization table 10, a high-compression rate quantization table 11, a code bit number detecting part 4, and a quantization table selecting part 12 are provided for quantization and encoding. The quantization threahold as the reference of quantization is stored in the table 10, and the table 11 forcibly sets the higher-order high frequency component of the two-dimensional orthogonal conversion coefficient to be 0. The detecting part 4 detects the number of bits of a code, which is obtained by encoding the quantization coefficient of a standard quantization table 3, with a block as the unit. When the table 3 is selected by the number of code bits of the block by the selecting part 12, compressed data is automatically decreased smaller than a certain quantity.</p> |
