A Method of Detecting and Modeling Mining-Induced Deformation Areas in Interferograms

The lack of attention to frequent or illegal mining activities that cause localized rapid surface subsidence can lead to significant accidents. Therefore, it is of great significance to identify and monitor such activities. The dense fringes generated by deformation gradients can affect the effectiveness and accuracy of InSAR processing. We proposed a new processing method that utilizes a two-dimensional mixed Gaussian model and mining area models to generate simulated interferograms to expand the dataset. Then we use YOLOv8 automatically detect suspected mining deformation areas from large-scale interferograms. Finally, by combining optimization algorithms with mining deformation models, we invert and model the deformation phase of the mining area, obtaining the modeled mining deformation phase results. This method can successfully detect the deformation areas and recover the deformation fringes, providing a foundation for improving the accuracy of subsequent InSAR processing.