配电网不停电作业中的行为识别技术分析  

Analysis of Behavior Recognition Technology in Uninterrupted Operation of Distribution Network

作　　者：王耀华 纪坤华 苏伟 WANG Yaohua;JI Kunhua;SU Wei(Shanghai Nan Electric Power(Group)Co.,Ltd.Non Power Outage Operation Branch,Shangh 201100,China)

机构地区：[1]上海市南电力(集团)有限公司不停电作业分公司,上海201100

出　　处：《电子技术（上海）》2024年第8期300-301,共2页Electronic Technology

摘　　要：阐述采用立体相机的被动测距方法,通过计算机程序接口传输图像的三维点云数据,借助基于深度学习的目标识别算法,精准识别不安全行为所对应的目标物,并应用NanoDet算法识别不安全行为。This paper expounds the passive ranging method using a stereo camera,which transmits three-dimensional point cloud data of images through computer program interfaces.With the help of deep learning based object recognition algorithms,it accurately identifies the target objects corresponding to unsafe behaviors and applies the NanoDet algorithm to identify unsafe behaviors.

关 键 词：三维点云数据 行为识别技术 识别算法 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]