Remote Sensing Image Classification Based on Log-Gabor Wavelet and Krawtchouk Moments

To further improve the accuracy of remote sensing image classification, a classification algorithm of remote sensing image based on Log-Gabor wavelet and Krawtchouk moments is proposed in this paper. Firstly, multi-direction and multi-resolution filtering is performed on a remote sensing image by Log-Gabor filter to extract texture features of the remote sensing image. Meanwhile Krawtchouk moments invariants of the remote sensing image are calculated to serve as edge shape features of the remote sensing image. Thus a complete feature vector is constructed with the texture features extracted by Log-Gabor wavelet. Finally the remote sensing image is classified according to the extracted feature vectors by supporting vector machine. The classification result of remote sensing image is obtained. A large number of experimental results show that, compared with three recent classification algorithms of remote sensing image such as the algorithm based on Gabor wavelet, the algorithm based on Log-Gabor wavelet and the algorithm based on Krawtchouk moments, the proposed algorithm has a significant improvement in the subjective visual effect and objective quantitative evaluation index such as classification accuracy. It is a kind of effective classification algorithm of remote sensing image.