基于深度学习的不安全行为管理技术

Unsafe Behavior Management Techniques Based on Deep Learning

建筑业施工环境复杂、多变,安全事故屡屡发生,事故率高居不下,传统管理模式已不足以及时全面地对现场安全隐患进行巡查、甄别和处理。亟需一种高效的信息化安全管理手段,实现安全管理模式和效能上的突破。为此,将AI算法与视频管理系统进行有机结合,针对目前现场安全管理存在的不足和需求,在已有智能识别技术基础上进行改进和创新,对基于深度学习算法不安全行为智能识别技术展开研究,以适应施工现场安全管理场景,提高现场安全管控水平、管理效率及成本。

The construction industry construction environment is complex and changeable,safety accidents occur frequently,and the accident rate is high.The traditional management mode is not enough to timely and comprehensively inspect,screen and deal with the potential safety hazards on the site.It is in urgent need of an efficient information security management means to achieve a breakthrough in safety management mode and efficiency.To this end,AI algorithms and video management system are organically combined,aiming at some problems of the on-site safety management and demand,the improvement and innovation are carried out on the bases of existing intelligent recognition technology,the research of unsafe behavior intelligent recognition technology are conducted based on the deep learning algorithm,so as to adapt to the construction site safety management scenarios,improve the level of onsite security control,management efficiency and management cost.