Change Detection from High-Resolution Remote Sensing Image Based on MSE Model

At present, most of the traditional change detection methods from high-resolution remote sensing image are based on a feature information, the information of multi-feature information cannot be extracted, so it is difficult to detect the complete information. In order to solve this problem, a change detection algorithm of high-resolution remote sensing image based on multiview spectral embedding is proposed in the present paper. Firstly, change image is obtained using traditional difference change detection method, and multi-feature information is extracted. The feature vector information is fused by a MSE model and the complete change information can be obtained. The experimental results show that the detection accuracy of the proposed method is better than the accuracy of traditional methods, and its stability is outstanding.