面向空间天文观测的序列图像无损压缩算法

空间天文观测任务会获得大量天文图像.对定点天体连续观测得到的序列天文图像具有时间及空间冗余较高的特点.为了减少序列天文图像的存储与传输的数据量，保证序列天文图像的完整性，满足科学目标的任务需求，需要对其进行无损压缩.本文提出了一种利用帧内压缩与改进的帧间压缩相结合的无损压缩算法，将序列天文图像的第一帧进行JPEG-LS帧内无损压缩编码，其余帧进行改进的帧间无损压缩编码，从而有效去除序列天文图像的时间及空间冗余，提高序列天文图像的压缩比.经过试验测试，改进后的帧间压缩效果优于帧内压缩效果，改进后的帧间压缩时间少于帧内压缩时间.结果表明，该算法简单且高效，适用于对序列天文图像的无损压缩. 

Sequential Imagery Lossless Compression Algorithm for Space Astronomical Observation

In astronomical observation missions, a large number of sequential imagery is obtained. They are produced by fixpoint photography in a period of time, and all have some common features with high resolution and high time-space redundancy. The large volume of such imagery challenges on-board transmission and storage mission. Hence, an algorithm based on inter-frame prediction is proposed in order to reduce the images' bitrates. It includes the intra-frame compression algorithm and an improved inter-frame compression algorithm. The JPEG-LS is adopted to encoding the first image as the intra-frame compression algorithm. Then the other images are compressed by the inter-frame compression algorithm. First, the next image is predicted by its previous image, and then the prediction residuals are transmitted to JPEG-LS predictor for refining. Finally, the refined residuals are encoded by arithmetic encoder. Arithmetic coding does not require encoding each pixel like Huffman coding, and it is closer to the entropy of the image. The experiment result proves that the proposed algorithm promotes the compression performance, and it has lower complexity and less compression time comparing with the traditional JPEG-LS. The proposed algorithm is suitable for real-time onboard compression of astronomical observation sequential imagery.