A coal mine law enforcement system based on multi-semantic scenario collaborative perception is developed, which includes mine safety production law enforcement platform, mobile law enforcement equipment and mine safety control cloud platform. Through the multi-modal fusion technology for multi-semantic scene collaborative perception, the multi-modal and multi situation decision-making model is constructed by fusing multiple information such as text, image, voice, video, spatial location and so on. Through information transformation and fusion, mobile law enforcement scene recognition and business automatic navigation can be realized. A closed-loop law enforcement link is constructed to automatically identify law enforcement scenarios through multimodal data fusion methods, and trigger law enforcement tasks, law enforcement processes, and law enforcement methods, so as to achieve precise law enforcement underground. Law enforcement data can be transmitted to the coal mine safety management and control cloud platform, which greatly improves the efficiency, objectivity of supervision and law enforcement.
