Environmental Restoration in Mining Areas: An Integrated Framework for Monitoring, Multi-Criteria Assessment, and Deep Learning Detection

This thesis develops practical ways to track how mining areas recover and how sustainable that recovery is. It uses satellite images to measure how vegetation changes over time in four mining regions in China, the United Kingdom and Malaysia. It then combines environmental, social and economic information to compare sustainability across sites. It also tests a deep-learning method to spot high-risk waste facilities, such as tailings ponds, from remote-sensing images. The work helps governments and companies monitor mining impacts, target rehabilitation efforts, and reduce safety and environmental risks.