A New Coeff-Token Decoding Method With Efficient Memory Access in H.264/AVC Video Coding Standard

In general, a large amount of memory accesses are required for context-based adaptive variable-length coding decoding in H.264/AVC standard. It is a serious problem for applications such as videophone and digital multimedia broadcasting services because of the considerable power consumption of the memory access function. In order to reduce the memory access, we carefully examined the codewords in variable-length code tables (VLCTs) of the coeff-token and found some distinctive features from the structure of codewords. Based on the features, a new coeff-token decoding method with an efficient memory organization is proposed in this paper. The simulation results show that the proposed algorithm achieves an approximately 95% memory access saving without video-quality degradation, compared with conventional decoding

I. Introduction

The appearance and development of various multimedia services have been based on high coding efficiency. The ITU-T/ISO/IEC Joint Video Team established the newest video coding standard known as H.264/AVC [1]. This standard achieves a considerable improvement over previous video coding standards. However, the new features and functionalities [2] adopted in the H.264/AVC result in an increase of the computational complexity not only in encoding but also in decoding. The complexity increment of the decoder gives rise to a cost-effectiveness problem of the development of an H.264/AVC-based video solution [3]. Thus, interest in implementing an H.264/AVC decoder has increased in recent years [3]–[5].