一种基于图象恢复和小波包分解的遥感图象压缩方法

An Algorithm of Remote Sensing Image Compression Using Image Restoring and Wavelet Analysis

设计了一种基于小波分解等去相关技术和图象恢复技术的遥感图象压缩＼解压方法。该方法首先分别利用比特分割和差值两种方法对图象进行去相关预处理，其次利用二层小波包分解将图象分解为低频和高频部分等１６个部分，然后对各高频成分进行了量化压缩或置零（去除），对低频部分采用无损的算术编码。为了提高解压重建图象的质量，利用线扩散函数对重建后的图象进行恢复处理。对ＳＰＯＴ和ＴＭ图象的实验结果表明，该技术是可行的。

An algorithm of remote sensing image compression, based on image restoration and wavelet bag analysis, is proposed in this paper. The pretreatment of bits segment and removing differential correlation are performed, and the image is decoded into 16 sub-bands with two level wavelet bag decomposing, which are respectively low and high frequency parts of image. Next, the high frequency parts are compressed by quantum method or seting to be zeros, and low frequency parts are coded using arithmetic coding. In order to improve the quality ofreconstruction image, a restoration technique based on line spread function is applied to it. The test of TM and SPOT remote sensing images shows that this algorithm is reasonable and applicable.