Estimating soil erosion using MODIS and TM images based on support vector machine and à trous wavelet |
| |
| Authors: | SH Chen HB SuJ Tian RH ZhangJ Xia |
| |
| Institution: | Key Laboratory of Water Cycle and Related Land Surface Processes, Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences, 11A, Datun Road, Chaoyang District, Beijing 100101, China |
| |
| Abstract: | To date, there is little work concerning the application of fusing images with significantly different spectral and spatial resolutions. In this paper, a novel method based on support vector machine (SVM) is proposed to quickly estimate soil erosion using the fused results produced from fusing such multisensor images by à trous wavelet transform (AWT). In the proposed method, the AWT is used to derive the high-resolution vegetation coverage image (HVCI) while the SVM overlays the HVCI and the slope image to derive the soil erosion map. By taking MODIS and TM images as an example, the potential of the proposed method is evaluated both quantitatively and qualitatively. The results show that it is feasible to perform the fusion of MODIS and TM images and the soil erosion map produced from the fused images by the proposed method can be achieved with an accuracy level comparable to that solely from the TM images. The merging of MODIS and TM images partly solves the constrains associated with the TM data availability which is caused by the lower revisit frequency and narrower spatial coverage. |
| |
| Keywords: | TM MODIS Support vector machine À trous wavelet transform Soil erosion |
| 本文献已被 ScienceDirect 等数据库收录！ |
|
