摘要
在铁路现场作业的过程中,人员的检测和识别一直是铁路视频监控系统的重点之一,而现有的检测算法存在检测速度慢、鲁棒性差、精度较低、实时性等问题。论文重点研究了基于深度学习的路服细粒度检测识别方法,利用该领域大规模数据集,以YOLO V2高性能检测算法为基础,引入混合注意力机制。实验表明该方法平均检测精度达到82.8%,且具有强鲁棒性和实时性,达到铁路现场监控要求。
In the process of railway field operation,the detection of personnel has been one of the key points of railway video monitoring system.However,the existing detection algorithms have some problems,such as slow detection speed,poor robustness,low precision,real-time and so on.Therefore,This paper focuses on the research of fine grain detection and identification technology based on depth learning.In this paper,a hybrid attention mechanism is introduced on the basis of YOLO’s high performance detection algorithm.Experimental results show that the algorithm achieves 82.8%detection accuracy,and is robust and real-time.It meets the expected effect and shows that the method is reasonable and feasible.
作者
刘家辉
胡广朋
王申宇
刘畅
覃源
程科
LIU Jiahui;HU Guangpeng;WANG Shenyu;LIU Chang;QIN Yuan;CHENG Ke(School of Computer Science,Jiangsu University of Science and Technology,Zhenjiang 212003;China Railway Tunnel Group No.3 Co.,Ltd.,Shenzhen 518051)
出处
《计算机与数字工程》
2020年第6期1367-1371,共5页
Computer & Digital Engineering