Research On Key Points Of H.264 Video Encoding System Based On FPGA

The H.264 video encoding system based on FPGA is a research focus on current video surveillance field, it is a good solution combining programmability and updatability of DSP scheme and the high programming performance of ASIC scheme ,having the advantages of easily upgrading the system architecture, flexible interface control, low-risk, high stability and high cost, etc. The traditional video encoding systems do not have such advantages.This paper studies several key issues in designing the H.264 video encoding system based on FPGA and presents corresponding solution methods based on the existing research and related tools in the laboratory, the main work is as following:(1) Completed the SoPC system building on the XUP Virtex-II Pro development platform, deeply analyzed the mechanisms of operating and constructing system, and transferred the Linux operating system, including establishing cross-compilation tools, generating BSP, configuration of Linux kernel, creating root file system and other related contents.(2) After high level analysis and comparison of two methods of entropy coding, CAVLC is selected to implement hardware encoder. This paper uses pipelined technology to achieve the design of CAVLC hardware encoder based on FPGA, and completes the synthesis and simulation of each module in encoder under the ISE.(3) Using XPS tool to design the CAVLC hardware encoder as a custom IP core, and modifies the relevant system files of the SoPC system, and then loads the CAVLC IP core on the SoPC system and tests it.(4) According to the related interface and encoding mechanisms of the custom CAVLC hardware encoder, develops corresponding drivers, ported them to Linux systems, and write an application to launch CAVLC hardware encoder, loading 4x4 block data to obtain the same encoding results with the theory data.