SVM classification of SAR images based on texture and internal edge
.基于纹理和边缘的SAR图像SVM分类*

For efficient land cover classification in SAR images,this paper studied 4 textural features based on GLCM and 2 edge features.First,it analyzed every individual feature for their discriminative ability for SVM SAR image classification into various land cover types.And then,chose several feature sets for SVM SAR image classification,and fused these features using PCA.The difference between PCA-and non-PCA-based classifications was also analyzed.The result shows the feature set comprising energy,edge length,...