摘要
为了准确、实时检测铁轨上方行人,避免出现严重的交通事故,提出基于BP神经网络的铁轨上方行人检测方法。采用双边滤波、伽马变换的方式增强铁轨交通监控视频图像,完成图像增强后,采用铁轨交通监控视频图像目标特征信息采集模型,提取增强后图像目标特征信息。将所提取的图像目标特征信息作为基于BP神经网络的行人检测器的输入信息,输出结果即为行人检测结果,完成铁轨上方行人检测。实验结果表明方法可有效优化铁轨交通监控视频图像质量,检测结果符合实际,且检测实时性显著,可应用在铁轨交通监控任务中。
Accurate and real-time detection of pedestrians above the railway track can avoid serious traffic accidents.Therefore,a pedestrian detection method above the railway track based on BP neural network is proposed.The rail transit monitoring video image is enhanced by bilateral filtering and gamma transform.After image enhancement,the target feature information of the enhanced image is extracted by using the rail transit monitoring video image target feature information acquisition model.The extracted image target feature information is used as the input information of pedestrian detector based on BP neural network,and the output result is the pedestrian detection result to complete the pedestrian detection above the railway track.The experimental results show that the proposed method can effectively optimize the quality of rail transit monitoring video image,the detection results are in line with the reality,and the detection real-time is remarkable.It can be applied to rail transit monitoring tasks.
作者
徐炳辉
苏有斌
白宏权
张岩
XU Bing-hui;SU You-bin;BAI Hong-quan;ZHANG Yan(Guoneng Shuohuang Railway Development Co.,Ltd.,Suning 062350 China)
出处
《自动化技术与应用》
2023年第5期76-79,共4页
Techniques of Automation and Applications
基金
国能朔黄铁路发展有限责任公司科技项目(SHTL-21-07)。
关键词
BP神经网络
铁轨行人检测
双边滤波
监控图像提取
BP neural network
rail pedestrian detection
bilateral filtering
Critical image extraction
