Optimal encoding of motion compensated video

摘要

The present invention involves a system and method for optimizing video encoding. For each candidate motion vector, encoding distortion is determined between a macroblock and a reconstructed macroblock by determining discrete cosine transform coefficients of the macroblock and quantizing the discrete cosine transform coefficients. An estimate unit estimates the length of the bit stream that would be required to encode the quantized discrete cosine transform coefficients along with the mode information bits including mode and motion vector information. The reconstructed macroblock is determined based on the quantized discrete cosine transform coefficients. A bit-rate term based on the length of the bit-rate stream is determined and included in the encoding distortion. The candidate motion vector which minimizes the encoding distortion of the macroblock is chosen to be the motion vector for the macroblock.