基于混合编码的天文序列图像无损压缩算法

提出了一种将时间和空间双维度预测与混合熵编码结合的图像无损压缩算法，应用于星体定点成像的天文序列图像.时间与空间双维度预测可以消除天文序列图像的时间相关性和空间相关性.混合熵编码融合了高熵编码与低熵编码算法，充分利用预测误差分布的稀疏性，进一步提升压缩效率.针对星体的天文序列图像，该算法首先在时间维度上进行帧间预测，去除时间相关性，再对帧间预测后的图像在空间维度上进行直方图滤波和上下文预测，去除空间相关性，最后将预测误差送入熵编码器进行编码.实验结果表明，本文所提出算法的压缩效果优于JPEG-LS，平均压缩比提升约15%.该算法结构简单，计算复杂度较低，易于硬件实现，适合星载天文图像的无损压缩. 

A Lossless Compression Algorithm for Astronomical Sequential Images Based on Hybrid Coder

An image lossless compression algorithm based on time and space dimension prediction and hybrid entropy encoder is proposed. The algorithm is applied to the astronomical sequential imagery produced by fixed-point photography. The time and space dimension prediction can eliminate the time correlation and spatial correlation of astronomical sequential imagery, and the hybrid entropy encoder combined with high-entropy coding and low-entropy coding makes full use of the sparsity of the prediction error distribution. For sequential imagery, the algorithm first performs inter-frame prediction on time dimension to remove the time correlation, and then adopts histogram filtering and context prediction on spatial dimension for the residual imagery of inter-frame prediction to remove the spatial correlation. Finally, the prediction error is sent to the hybrid entropy encoder for coding. The experimental results show that the performance of the proposed algorithm is better than that of JPEG-LS, with an average compression ratio of about 15% improvement. The algorithm has simple structure and low computational complexity, which is friendly for hardware implementation and suitable for lossless compression of space astronomical imagery.