Abstract: In view of problems of complex devices and needing a large amount of water in current common raw coal separation methods, an online separation system for coal and gangue based on dual-energy γ-ray and digital image recognition technology was designed. The single-chip microcomputer real-timely judges materials under detector whether coal or gangue, if it is coal, spray gun has no-action, if it is gangue, the single-chip microcomputer controls the spray gun spray high-pressure gas to insufflate the gangue into gangue channel. When the coal has many impurities or the gangue has many carbons, image recognition module distinguishes coal and gangue according to gray differences between coal and gangue, and sends recognition result to the single-chip microcomputer. The simulation test result shows that the system can recognize coal and gangue effectively and its recognition precision meets requirements of industrial production.