靶场光测图像序列的复合压缩方法

针对靶场海量测量图像的存储问题，提出了光测图像序列的复合压缩方法。该算法解决了小波变换产生浮点数，导致信号不能精确重构的问题，实现了从整数变换系数中完全重建图像；同时借鉴ROI思想，解决了图像压缩需求与测量精度的矛盾，在同一压缩框架下，实现了目标区域的无损压缩和背景区域的高倍压缩；设计了靶场测量图像序列压缩码流结构，保证了解码时目标区域信息的完整性、目标区域和背景区域压缩码流的可分割性和整体码流的连续性。经仿真验证，该算法科学可行，在同一压缩比时本文方法的SSIM值优于JPEG2000压缩标准，压缩后目标判读误差在3”内，在满足数据处理精度的基础上实现了光测图像的大比例压缩。

Composite compression method for range optical measurement image sequence

To meet requirements of storage of the mass optical measurement image in range, a composite compression method for range optical measurement image sequence is proposed in this paper. By solving the problem that signal can not be reconstructed precisely because of the floating-point coefficients produced by the wavelet transform, this algorithm can reconstructed the image completely from the integer transform coefficients; at the same time, the contradiction be- tween the compression ratio and precision is solved by the thought of ROI which using lossless compression for the target area and high-powered compression for the background area in the same framework. The structure of the compressed code stream of the range measurement image sequence is well designed, in order to protect the integrity of the target area for information and the divisible and continuity of the hole code stream. The simulation results show that, this method is sci- entific and feasible, the SS｜M values of it is better than that of JPEG2000 at the same compression ratio, the parallax error is less than 3 seconds of arc, and the image compressed by it even in a large ratio can also meet the requirements of the da- ta processing.