摘要
人脸识别技术是提取及比对脸部特征信息进而鉴别身份的生物识别技术。针对电力巡检机器人视频监控中的身份鉴别问题,本文采用深度学习技术中最前沿的卷积神经网络算法作为人脸识别的核心算法,重点研究卷积神经网络在人脸识别中的实现过程,进而完成实验室的姿态变化人脸检测性能测试以及现场环境下的电力巡检机器人实时人脸识别测试。试验结果显示卷积神经网络算法可以准确完成身份鉴别,极大地提高了变电站中电力巡检机器人的人脸识别能力与安防能力。
Face recognition technology is a biometric technology that extracts and compares facial feature information to identify the identity. For the identification of power patrol robots in video surveillance,this paper uses the most advanced convolutional neural network algorithm in deep learning technology as the core algorithm of face recognition,focusing on the realization process of convolutional neural network in face recognition. Then complete the change of the posture of the face detection performance test and real-time face recognition test of the power inspection robot in the field environment. The test results show that the convolutional neural network algorithm can accurately complete the identity authentication,which greatly improves the face recognition ability and security capability of the power inspection robot in the substation.
作者
马力
王致
张丹
洪永健
王天安
MA Li;WANG Zhi;ZHANG Dan;HONG Yongjian;WANG Tianan(Yunnan Power Grid Co.,Ltd.Kunming Power Supply Bureau,kunming Yunnan,650011)
出处
《自动化与仪器仪表》
2019年第2期36-38,共3页
Automation & Instrumentation
关键词
人脸识别
卷积神经网络
巡检机器人
安防
face recognition
convolutional neural network
inspection robot
security
