Support Vector Machine for Classification of Hyperspectral Remote Sensing Imagery

As one of the popular and advanced statistical learning algorithms, support vector machine (SVM) has been the new hot study area of pattern recognition and machine learning in recent years. SVM has such advantages as suitableness to high dimensional data, requirement of few samples and robustness to uncertainty, so it can be used to hyperspectral remote sensing image classification effectively. Based on the theory of SVM, a new approach for information classification on hyperspectral sensor has been developed by the experimental case of spatial information classification in central area of Shanghai city with PHI image. The algorithm is synthetically compared with the traditional classification methods. The experiment results confirm the effectiveness of the proposed method, which results in higher classification accuracy than the traditional methods.