Manner of quantization estimating the data compression device and quantization estimate circuit null of data compression

摘要

PURPOSE:To provide a quantization estimate method in data compression and a quantization estimate circuit in which a fixed area is utilized effectively to attain data processing with high quality. CONSTITUTION:A code length for 1 DCT period is calculated for each of M-sets of quantizers (i), a code length of each macro block comprising plural DCTs is accumulated sequentially for each quantizer, a quantizer used in common is decided in the unit of macro block slice comprising plural macro blocks in a degree not causing overflow, and a total code length is calculated when a quantization level (qi-1) of the decided quantizer is replaced with a higher quantization level qi for each macro block sequentially and optimum quantizers Qi, Qi-1 usable in an extent of not causing overflow are decided in the unit of macro blocks. In the case of accumulation, the half-way result and the final result of the accumulation of code length for each small block are stored in N-sets of staircase memories and the total code length is calculated when the quantization level is replaced by utilizing it.