基于SPOT 5图像的岩溶地貌单元自动提取方法

通过对峰林、峰丛和岩溶洼地3者的地理特征和影像特征的研究,基于遥感图像本底值提出了能有效反映目标特征的遥感指数——植被指数、土壤亮度指数、图像主成分变换第1主成分值及地形数据等,并构建了遥感指数的集成计算法,建立了遥感自动提取模型.指数集成运算法能够有效地增大峰丛、峰林与其他地物之间的光谱差异,使这些岩溶地貌单元的灰度值高于其他地物,从而利于岩溶地貌单元提取阈值的自动选取.基于构建的遥感自动提取模型先提取了峰丛、峰林信息,并在此基础上提取了岩溶洼地信息.经实验研究表明,该方法具有较高的提取精度和效率.

Automatic Extraction of Karst Landscape Elements Based on SPOT 5 Image

To study the geographical features and image features of peak-cluster,peak-forest and karst depression,this paper puts forward some remote sensing indices based on background values of remote sensing images,such as vegetation index,soil brightness index,and PC1 of principal component transformation of the image and terrain data.Meanwhile,the integrated calculation method of remote sensing indices is proposed,and the automatic extraction model of remote sensing is created.The integrated calculation method of the indices could effectively increase spectral differences between peak-cluster,peak-forest and other surface features.The gray values of peak-cluster and peak-forest are the highest in the image so as to obtain segmenting value for accurate extraction of them based on automatic selection algorithm of threshold.Based on automatic extraction model of remote sensing,this paper puts forward some information of peak-cluster and peak-forest.On such a basis,karst depression information is extracted.Experimental studies show that the method has high accuracy and efficiency of extraction.