深度学习在口岸场所视频智能监管中的框架设计

Framework Design of Videos Intelligent Supervision of Port Sites with Deep Learning

本项目利用深度学习对口岸场所视频监控采集数据流进行目标识别、分离,行为分析、标定,规则判定、预警等,分析从周界警戒、堆场运提、仓储管理、检查验核、卡口监控等各环节相关的目标及行为,从而满足不同监管要求的智能管理与决策,最终实现口岸场所全封闭、全天候作业,监管部门全链条、全方位监管。

This project uses deep learning to conduct object identification,separation,behavior analysis,calibration,rule determination,early warning,etc.on the video surveillance data stream of the port sites,and analyze the objects and behaviors of the perimeter alert,container-yard transportation,storage management,inspection and verification,gate monitoring and other related processes,so as to meet the intelligent management and decision-making of different regulatory requirements.Finally,the department shall supervise the whole chain and all directions in port sites,while them are in the closed and all-weather operation.