Research And Implementation Of Fpga-based Lossless Video Compression Algorithms

Certain applications such as medical video and aerial video arching, where videos are sensitive to distortion, require lossless compression. As FPGA is working parallelly, the efficiency will be high when it is used to process video. The research on FPGA-based lossless video compression is not much so far. So research and implementation of FPGA-based lossless video compression algorithms have important theoretical value and practical significance.Existing lossless video compression algorithms are either low in compression ratio or too complicated to implement with FPGA. This thesis proposes an adaptive lossless video compression algorithm based on video flatness after full investigation of traditional lossless image compression algorithms. Work and achivements are presented as follows.(1) Workflows of CALIC, JPEG-LS and JPEG2000are elaborated first. Then a series of experiments on compression time and compression ratio are conducted to find which algorithm is better considering both complexity and compression ratio. Finally a lossless video compression algorithm based on flatness is proposed. The algorithm combines JPEG-LS, frame difference with image flatness and experiments results show that the proposed algrithm is effective for those videos not moving violently.(2) Hardware architecture of lossless video compression system is proposed according to requirement analysis. The system is able to encode or decode2-channel video at the same time.(3) Implementation scheme of the proposed algorithm is presented. Module HDMI in and module controller of DDR2SDRAM are connected by VFBC interface while video encoding and operation of SD card are completed by MicroBlaze. The SD card is operated in SPI mode and configured with FatFs file system.