Building change detection method based on differential visual field fusion and collaborative awareness

The technical problem to be solved by the invention is to provide a building change detection method based on differential visual field fusion and collaborative perception, a twin network is used as an encoder, a differential visual field fusion module is designed, and the method aims to more comprehensively capture building change information and reduce picture information loss caused by scale change at the same time. Boundary information blur can be processed more effectively; meanwhile, a designed collaborative perception module is also applied to capture the long dependency relationship and position information of a high-dimensional image, and the context between pixels is modeled, so that the integrity problem of a changing object is effectively solved. On the basis of the two modules, a difference technology is applied in a decoder, so that the image segmentation capability of the model is improved, and a better detection effect is obtained. 本发明所要解决的技术问题是提供一种基于差分视野融合与协同感知的建筑变化检测方法，使用孪生网络作为编码器，设计一种差分视野融合