Landslide Recognition in Remote Sensing Image Based on Fuzzy Support Vector Machine

Landslide is one of the most important natural disasters, which has wide distribution region, high frequencies of occurrence and fast movement speed. The landslide data have characteristic of highly nonlinear, fuzzy features and a large amount of data. In this paper, considering the characteristics of the landslide data and the shortage of SVM, the fuzzy support vector machine with textural features is introduced to identify landslide in remote sense image. By improving the fuzzy membership to overcome the influence of noise to the training process and improving the penalty coefficient to eliminate the negative impact of un-balanced sample size, the accuracy of the landslide recognition is further enhanced. Finally, the information of landslide can be extracted by using the remote sensing images of the disaster area. Using fuzzy support vector machines to extract the landslide is effectiveness and feasibility in remote sensing images, which is proved by instances.