Abstract:

With the rapid development of unmanned intelligence and deep learning technology, the traditional manual detection method has become increasingly inefficient and limited in the detection of surface unexploded substitutions. In response to the problems of slow ground detection speed and large errors in aerial detection, this paper proposes a rapid detection system based on an unmanned aerial vehicle (UAV) platform. The system adopts a multi-mode imaging detection, artificial intelligence target detection, and unmanned load inspection approach. Through the collaborative work of three subsystems: the UAV-mounted imaging detection platform, the two-dimensional map rapid detection system, and the ground station, it achieves all-day efficient detection of surface unexploded submunitions. Experimental results show that the system significantly improves detection speed, accuracy, and environmental adaptability compared to other traditional detection methods. Moreover, it does not require manual close contact, effectively reducing safety risks, and provides a new solution for the detection of surface unexploded submunitions.

Key words: surface unexploded submunition, unmanned aerial vehicle (UAV), deep learning, rapid detection, two-dimensional map reconstruction