煤矿智能视频预警系统架构及应用场景研究

Architecture and application scenario of intelligent video early warning system in coal mine

为提升煤矿安全风险分级管控和事故预测预警能力,实现煤矿行业企业的动态监管和风险防控,基于人工智能和机器学习等视频分析技术,实现对井上、井下诸多工作场景进行智能监控,自动发现隐患并报警,建立了煤矿智能视频预警系统的开发平台架构、业务应用架构,提出了煤矿智能视频预警分析指标体系、预警防控与分析方法、建立安全生产统一信息平台;分析了以视频图像数据为依托,GIS地图动态全景展示工作区规范佩戴安全帽、轨道站人识别、趴蹬运输车行为识别、重点岗位人员脱岗识别、胶带跑偏检测、胶带空载检测、井下烟火检测等7个智能视频预警系统应用场景。

In order to improve the hierarchical management and control of coal mine safety risks, enhance the ability of accident prediction and early warning, realize the dynamic supervision and risk prevention and control of enterprises in the coal mine industry, the development platform architecture and business application architecture of the coal mine intelligent video early warning system was established using video analysis technologies such as artificial intelligence and machine learning to intelligently monitor the surface and underground working scenes, automatically detect hidden dangers and alarm. The index system of intelligent video early warning analysis of coal mines, early warning prevention and control and analysis methods were was proposed and a unified information platform for safety production was established. Based on video image data, seven application scenarios of intelligent video early warning system were analyzed, including GIS map dynamic panoramic display of working area, standardized wearing of safety helmets, identification of railway station personnel, identification of crawling and pedaling transport vehicle behavior, identification of key post personnel off post, belt deviation detection, belt empty load detection and underground fireworks detection.