Comparison of context-based adaptive binary arithmetic coders in video compression

A comparison of two context-modeling methods, CABAC and GRASP, is presented with regard to applications in video compression. Both context-based entropy coding methods are tested for various probability estimation techniques, for different sources, and for different quantization parameters by using the test model of the H.264/AVC video coding standard. Our experimental results show that GRASP provides a significant gain over CABAC, especially for intra frames coded at medium to high bit-rates.