Implementation of an Efficient and Low Cost CAVLC for MPEG-4 AVC/H.264 Video Coding Application

• Main Authors: Shu-Ping Chang, 張紓萍
• Other Authors: Tsung-Han Tsai
• Format: Others
• Language: en_US
• Published: 2008
• Online Access: http://ndltd.ncl.edu.tw/handle/25150236738093876610

碩士 === 國立中央大學 === 電機工程研究所 === 96 === In order to achieve the higher compression ratio, Context-Based Adaptive Variable Length Coding (CAVLC) is adopted as one of entropy encoder in MPEG-4 AVC/H.264 [1], [2]. Compared with the traditional entropy encoder, CAVLC can achieve better coding efficiency, but the algorithm complexity is higher. On the other hand, because of the data dependency in CAVLC, it results a complex CAVLC encoding in hardware implementation. At the same time, throughput is the other concern in CAVLC, especially for higher resolution video such as HDTV. Deal with higher resolution of video information, encoding (or decoding) with an efficient CAVLC encoder is important. In this paper, an efficient and low cost CAVLC encoder is proposed for video coding applications of MPEG-4 AVC/H.264. Two main concepts are proposed to increase the throughput. One is the combination of scan phase and coding phase; the other is the block-based pipelining by the associated input buffer. We also propose three schemes to speed up the coding efficiency as well as reduce the area cost. One is to simplify the seven LEV-VLCN tables used by level coding. Another is to generate the nC value according to six block modes. The last one is to reduce the memory size used by nC generator. And the proposed CAVLC design is downloaded to FPGA (Xilinx Virtex-4 XC4VLX60), and co-simulation with the h.264 source JM10.1 encoder c program for further verification. The encoded bitstream file (bitstream.264) can be decoded by h.264 source JM10.1 decoder c program and played successfully.