多融合的入侵检测系统在电站安防应用  

Application of multi-fusion intrusion detection system in power plant security

作　　者：郭义明 郁启华 邵书成 Guo Yiming;Yu Qihua;Shao Shuchen(SPIC GUANGXI ELECTRIC POWER CO.,LTD,Nanning Guangxi,530000;CHN ENERGY ZHISHEN CONEROL TECHNOLOGY CO.,LTD,Beijing,100000)

机构地区：[1]国家能源集团广西电力有限公司,广西南宁530000 [2]国能智深控制技术有限公司,北京100000

出　　处：《电子测试》2022年第7期116-118,共3页Electronic Test

摘　　要：为了提高电站中视频监控的智能化检测,提高电站无人化值守的推进。本文针对现有电站的监控设备进行功能改进,从视频图像移动检测、视频画面目标检测、视频画面人脸检测和识别方面进行算法功能融合,开发智能一体化无人值守乳入侵检测系统。本系统采用跟踪算法能够适用于多路并发的视频接入的电站运维,有效的解决监控视频中,入侵人员发现的滞后性,提高了电站运维的效率。In order to improve the intelligent detection of video surveillance in power station,improve the promotion of unmanned power station on duty.This paper improves the functions of the monitoring equipment of the existing power station,integrates the algorithm functions from the aspects of video image movement detection,video picture target detection,video picture face detection and recognition,and develops an intelligent integrated unattended milk intrusion detection system.The tracking algorithm adopted in this system can be applied to the operation and maintenance of power station with multi-channel and concurrent video access,which can effectively solve the lag found by intruders in the surveillance video and improve the efficiency of power station operation and maintenance.

关 键 词：电站入侵检测 人工智能检测 目标检测 卷积神经 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]