Data compression using adaptive bit allocation and hybrid lossless entropy encoding

(57) The method and equipment of summary adaptation bit assignment and hybrid reversible entropy coding. This system contains three components, (1) conversion stage, (2) quantization stages, and (3) reversible entropy encoder stage. A wavelet transform algorithm is used for the conversion stage (1). The quantization stage (2) presumes the value for the parameter which defines approximation between quantization size and the logarithm of quantization errors accommodative, and calculates the quantization size optimal for each band recursively that the desired bit rate should be attained. Baseband and a subband are changed into a quantization matrix using corresponding quantization size. The reversible entropy encoder stage (3) uses the consideration fact that the entropy characteristic of the run length of the zero index value in a subband quantization matrix differs from the entropy characteristic of a non-zero index. The syntax of each quantization matrix is analyzed so that each non-zero index may be extracted in a separate stream, and about "0", the syntax of the remaining position information is analyzed to the odd number stream of a run length value, and those of it is analyzed about "1" to the even number stream of a run length value. Huffman encoding of these three streams is separately carried out in the conventional way.