Object change detection and classification with a binary change-guided model using remote sensing images

Object change detection of remote sensing images has become a research hotspot in recent years. It is currently widely applied in areas such as land resource surveys, dynamic monitoring and geological disaster investigation. Compared to traditional methods, remote sensing image object change detection and classification methods based on deep learning have achieved significant improvements in accuracy. Despite the great success of deep learning-based feature extraction methods, their supervised learning style still has shortcomings and limitations, such as labeling method of remote sensing data and limited samples. Addressing these challenges, this paper conducts research on object change detection and classification with a binary change-guided framework. The proposed model consists of three parts: a change detection network,which distinguishes between changed and unchanged categories. Image segmentation segment the dualtemporal images to extract geographic objects. The classification network determines the type of change. Experimental validation shows that this model can effectively extract the feature information in different categories, achieving outstanding object change detection and classification results in remote sensing images.