摘要
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.
作者
余泽辉
黄传恭
王震
何露华
YU Zehui;HUANG Chuangong;WANG Zhen;HE Luhua(Fuzhou Customs Information Center,Fuzhou Fujian 350015)
出处
《软件》
2023年第7期57-60,共4页
Software
基金
福建省科技计划项目(2021R0112)。
关键词
深度学习
场所视频
智能监管
deep learning
venue video
intelligent supervision