VHDL Modeling of an H.264/AVC Video Decoder
摘要
Transmission and storage of video data has necessitated the development of video com pression techniques. One of today's most widely used video compression techniques is the MPEG-2 standard, which is over ten years old. A task force sponsored by the same groups that developed MPEG-2 has recently finished defining a new standard that is meant to replace MPEG-2 for future video compression applications. This standard, H.264/AVC, uses significantly improved compression techniques. It is capable of providing similar pic ture quality at bit rates of 30-70% less than MPEG-2, depending on the particular video sequence and application [20].\nThis thesis developed a complete VHDL behavioral model of a video decoder imple menting the Baseline Profile of the H.264/AVC standard. The decoder was verified using a testing environment for comparison with reference software results. Development of a synthesizable hardware description was also shown for two components of the video de coder: the transform unit and the deblocking filter. This demonstrated how a complete video decoder could be developed one module at a time with individual module verifica tion. Analysis was also done to estimate the performance and hardware requirements for a complete implementation on an FPGA device.