Remote sensing technology has evolved from an auxiliary tool to a core technology for breakthrough discoveries in geological mineral exploration. This paper systematically reviews research progress in this field over the past five years. Hyperspectral remote sensing, with dozens to hundreds of narrow bands, enables precise identification of alteration minerals and boasts a detection depth of 500-1000 meters, tripling exploration efficiency compared to traditional methods. The integration of full-band hyperspectral sensors and magnetometers on Unmanned Aerial Vehicle (UAV) platforms overcomes complex terrain limitations, establishing an air-ground collaborative system of "satellite macro-scanning UAV target area fine identification ground verification," reducing the field verification cycle by 60%. Artificial intelligence algorithms are deeply involved throughout the exploration process; Random Forest (RF) and 3D Convolutional Neural Networks (3D-CNNs) improve mineral identification accuracy to 85% in multi-source heterogeneous data fusion, while reinforcement learning models optimize drill target layout, reducing costs by 40%. Empirical evidence shows that after applying intelligent remote sensing technology in Kazakhstan, exploration costs dropped from USD 1200 per square kilometer to USD 400, and the discovery cycle for new deposits was compressed by 65%. Current challenges include barriers to multi-source data standardization, insufficient generalization ability of AI models, and real-time bottlenecks in edge computing. The future focuses on the "green exploration + intelligent perception" paradigm, leveraging solar-powered UAVs and biogeochemical collaborative detection technology to advance exploration towards zero-carbon and intelligent transformation.
Di Hu (Tue,) studied this question.