Underground coal mining results in large goafs and numerous abandoned mines that contain substantial amounts of coalbed methane. If this methane is not used and controlled, it will escape into the atmosphere through geological fractures and can result in serious greenhouse gas effects and environmental damage. Exploring and developing the coalbed methane resources of abandoned mines can not only improve coal mine safety and protect the ecological environment but also reuse waste and mitigate energy shortages. Geophysical methods have made some progress in detecting abandoned coal mines, but there are still some challenges and difficulties. The resolution of seismic exploration may not be enough to accurately describe the details of coal seams and CBM rich areas, and the effect of resistivity method in deep CBM exploration is limited. In addition, the geological structure of abandoned coal mines is usually more complex, such as faults, folds, etc., which makes the application of exploration methods more difficult and increases the difficulty of data interpretation. Therefore, it is necessary to develop and perfect exploration technology continuously including the application of geophysical big data, deep learning, and artificial intelligence inversion to realize the accurate detection and evaluation of CBM resources in abandoned coal mines.