四叉树分类的网格编码量化在SAR图像中的应用

SAR image compression using TCQ based on quadtree classification

综合了小波去相干斑噪声、四叉树分类、网格编码量化技术,提出了一种在小波域内对带噪SAR图像作网格编码量化的新方法。首先对小波域的SAR图像实施软阈值去噪声,然后根据小波图像中各子带系数的带间相关性对其进行四叉树分类,再对分类后的重要类小波系数应用网格编码量化形成有序的嵌入式比特流。该方法不仅利用了信号间的时间相关性,而且也较好地利用了信号变换域的相关性。并且在对SAR图像压缩的同时去除了噪声,取得了很好的效果。

By combining with speckle noise reduction, quadtree classification with trellis coded quantization(TCQ) in the wavelet framework, a new method which performs TCQ of SAR image in wavelet domain based on quadtree classification is proposed. First, the wavelet domain soft-thresholding technique is implemented to reduce the speckle noise before compression is perfomed. And then, the quadtree classification utilizes correlation between the subbands. At last, TCQ is applied in list of significant pixels(LSP), ordering embedded output bit stream is formed. This method utilizes not only time correlation, but also the correlation in transform domain.Image compression is combined with speckle reduction. Simulation states this method has advantages in SAR image compression.