机载图像无损/近无损压缩方案及其FPGA实现

根据机载图像压缩和传输的特点,在改进JPEG-LS算法基础上,设计了一种有效的图像无损/近无损压缩方案.该方案由去相关处理、熵编码及压缩位率控制三部分组成.去相关处理部分解决了JPEG-LS预测模型本身的误码扩散问题,熵编码部分使用快速有效的Golomb熵编码器完成对预测误差的编码,同时解决了机载应用中数据传输率恒定条件下压缩码率的控制问题.对压缩算法现场可编程门阵列(FPGA)设计中的一些关键问题也给出了有效的解决途径,从而形成了一套完整、可行的机载图像压缩解决方案.最后通过FPGA实现和验证了压缩方案及逻辑设计的正确性和可行性. 

Scheme design and FPGA implementation of airborne image lossless and near-lossless compression

For plane on-board image lossless and near-lossless compression,an improved compression algorithm based on the JPEG-LS was given.The algorithm is composed of three parts,decorrelation processor,adaptive entropy coder and controlling of compression ratio.The new decorrelation model solved the problem of transmission error propagation in JPEG-LS.A fast and effective entropy coder named Golomb was modified and used in the algorithm,and that a valid controlling strategy of compression ratio was present with invariable speed of image data input and output,so a full and feasible scheme was given and simulated with VHDL for airborne image coding and transmission.Furthermore,the whole scheme was implemented and verified with one chip of field-programmable gate arrays(FPGA) manufactured by Xilinx Corp.Results show the encoder meets the application of airborne image coding well in bulk,weight,and power consumption,and all key techniques in FPGA design are referred to.