基于树型小波变换和尺度共生矩阵的居民地提取

利用基于树型框架小波变换和尺度共生矩阵的方法对遥感影像中居民地进行识别提取。通过非完全树型框架小波变换把图像变换到不同的尺度层上，然后在不同尺度层之间，提取相互依存的尺度共生矩阵信息，进而形成居民地的纹理特征。实验结果表明，用基于小波的尺度共生矩阵的方法对居民地有较好的提取结果。

Residential Area Extraction Based on Tree-structure Wavelet and Scale-based Concurrent Matrix

The paper distinguished residential area from remote sensing images based on the methods of tree-structure wavelet transforms and scale-based concurrent matrix. It transformed the image into different scales using tree-structure wavelet transforms. And then, it distinguished information on scale-based concurrent matrix. From these, the texture characteristics of residential area were formed. The result of experiment showed that： it could get a better distinguished result by way of scale-based concurrent matrix which was based on the wavelet transforms.