Research on Remote Sensing Image Classification Method Based on SVM & Clustering

Remote sensing image segmentation is important in the analysis of remote sensing images geomorphic features and spatial relationships. However, due to remote sensing image with complex data characteristics, the traditional methods used in remote sensing image segmentation cannot meet the requirements. This paper presents a method of remote sensing image segmentation method based on SVM. & cluster. Remote sensing image is processed mainly by way of human-computer interaction: training samples was selected and trained, then the support vector was obtained, and next each pixel of the image was classified. As for image possible spots/holes after classification, the method of clustering was adopted to remove. To verify the effectiveness of the algorithm, a map on Google Maps was selected, experiment results showed that this method is effective.