A method of rate control of wavelet picture coding using dynamic programming

The JPEG2000 method [ISO/IEC FCD15444-1, July 2002] [D. S. Taubman, July 2000], applied to compress still images or moving pictures frame-by-frame, performs bitplane arithmetic entropy coding on each coefficient block of the wavelet transform of the source images. The rate control problem is considered as a combinatorial optimization problem (the resource allocation problem) to decide the truncation bitplane level of each code block, which minimizes the summation of the quantization error over the whole image according to the total bitrate available for the compressed image. In this paper, a new method is proposed to decide the truncation bitplanes using a dynamic programming scheme. The algorithm and an experimental result are also reported.