Research And Development Of A FPGA-Based Image Acquisition And Compression System Of Road Environment

Video data and image data collected by a video-based mutifucntional inspection system of comprehensive information of roadway are usually prodigious, and they need to be compressed rapidly and efficiently for the convenience of data transmission and storage. Besides the high cost and poor extensibility,the performance of traditional image compression and processing systems based on industrial computers are limit to the CPU speeds and bus bandwidths. Thus how to speed-up the system and improve the compression efficiency through hardware compression in use of FPGA devices is the primary problem to solve in the thesis.First, the thesis reviewed image acquisition and compression techniques and FPGA techniques at home and abrod, and designed a high performance-price ratio JPEG compressing and encoding scheme based on FPGA.After the detailed user requirement analyses. According to the structure characteristics of Virtex-II PRO XC2VP30, a video decoding circuit that meets the requirements is designed and the JPEG compression algorithm is implemented on FPGA. The procedure is divided into three steps:(1) using the concurrent work of multiple multiplier, Discrete Cosine Transform(DCT) is realized with serial input and parallel output; (2) zigzag scan is integrated into the quantification procedure, and all the divisions are converted into efficient multiplications to reduce time; (3) according to the features of encoding coeefficients, suitable methods are applied to the encoding procedures of DC coefficients and AC coefficients of images respectively. Finally, a performance comparison of the developped FPGA-based compression system and the PC-based compression system is carried out.Experiment results show that for FPGA-based system, its compression quality is equivalent to that of PC, and superior to PC-based system in time and ratio of compression, andcan meet the needs of practical application.