This article proposes a Tansformer based change detection algorithm for hyperspectral remote sensing images.The algorithm applies PCA dimensionality reduction technology,uses 3D convolution and 2D convolution to extract deep local features,and combines the CBAM module and Tansformer encoder module to further utilize the correlation between the spatial and channel dimensions of hyperspectral images.The results show that the change detection algorithm proposed by this algorithm has high detection accuracy,strong timeliness,and can achieve a good balance between accuracy and complexity.
维普
万方数据