Data

Regional landslide extraction is the basis of landslide hazard risk management. Acquiring landslide location and size information timely and effectively is crucial for disaster prevention and reduction. However, existing methods face challenges in accuratelyidentifyinglandslides due to inadequate learning of object features. Moreover, previous studies on the automated extraction of landslides have predominantly focused on those triggered by seismic events and precipitation, mining-induced landslides extractionremains comparatively limited. Here, we develop the MRSD-U-Net model, an enhanced deep learning network integrating multi-source remote sensing data and residual learning module.This model enhances the extraction of landslide features by deepening the network and capturing topographic, textural, and spectral information. The performance assessment show thatMRSD-U-Netachieves lower error rates, higher accuracy, and stronger generalization ability compared to other advanced deep learning models, thereby enabling a more accurate extraction of landslide boundaries. Furthermore, the landslide areas extracted by MRSD-U-Net are the most consistent with ground reference data. The research providesan effective method for the rapid and precise extraction of mining-induced landslides.