Application research on binocular vision systems for unmanned aerial vehicles in underground coal mines

With the rapid development of smart mine construction, the demand for automation and intelligence in underground coal mine inspection has become increasingly prominent. Traditional manual inspection methods not only involve high labor intensity but also carry the risk of human error. To address this, this study innovatively adopts a solution featuring a quadcopter drone equipped with a binocular vision system. This system builds a real-time three-dimensional model of the underground environment through stereo vision, combines intelligent flight control algorithms to achieve autonomous obstacle avoidance, and transmits tunnel images to the ground in real time via a wireless communication link. Experimental verification shows that this solution maintains a safety pass rate of over 90% in underground environments without GPS signals, significantly enhancing the safety and reliability of mine inspection operations. This research achievement provides a practical and feasible technical path for unmanned inspection in underground coal mines.