基于图像分类方法滑坡识别与特征提取--以三峡库区归州老城滑坡为例

Resent years, stereoscopic images interpretation and compute-human interactive interpretation for landslide hazards are extensive methods in landslides hazards detection, which recognize landslides according to the diagnostic characteristics include shape, tone, texture, terrain, geomorphology, et al. But the methods did not make full use of the important potential (spectral) information hidden in images. On the basis of images interpretation for landslides in the Three Gorges Reservoir area, this paper make a case study on the spectral characteristics extraction and analysis. And the Guizhoulaocheng landslide was chosen for classification processed by ENVI. We conclude that image classification function can remove some useless information and help to landslide recognize and characteristics extraction by classifying the images into different classes we need.