SVM Classification for Sensitive Land Parcels in Highway Route Selecting Based on ALOS Image Combining NDVI

In order to improve the efficiency and cost-saving investigation for sensitive land parcels for road route selecting, this paper demonstrates the methodology of Support Vector Machine (SVM) classification combining with the Normalized Difference Vegetation Index (NDVI) to identify the land parcels using ALOS remote sensing data. One part of the road corridor is taken as the study area in City Group of Changsha, Zhuzhou and Xiangtan, which is regarded as the two society pilot area. The results show that the high effectiveness and applicability of the method in high density vegetation coverage mountainous regions.