摘要
路面交通线破损图像智能检测的研究极大的提高了对路面维护和保养的效率。针对现有系统获取图像较难且像素差、处理图像算法效率低等问题,提出了城市智能监控系统路面交通线破损图像智能检测算法。即通过城市智能监控系统获取图片样本,采用高斯低通滤波和脉冲耦合神经网络相结合算法对样本进行滤波分割处理来突出路面交通线等有效信息,智能检测和计算分析图片纹理特征并得出结果。根据标准样本和检测样本的检测结果进行对比分析做出结果判断,在检测出路面交通线存在明显破损时对市政部门提供及时的预警。实验结果表明,改进系统获取图片方便且像素高,同时检测速度快、可靠性高,可运用于实际路面交通线破损等路面状况的检测系统。
intelligent detection of damaged image for road traffic line can greatly improve the efficiency of pavement preventive maintenance.An intelligent detection algorithm of damaged image for road pavement traffic line based on city intelligent monitoring system is presented.The image samples are obtained through the urban intelligent monitoring system,filtered and segmented by combining the Gaussian low-pass filter with PCNN to highlight the effective information such as road traffic line,and the image texture features are obtained through intelligent detection and calculate analysis.The test results are judged by comparing the standard sample with the testing sample.If there are obvious damages in the detection of the road traffic line,then a warning is timely provided to the municipal administrative departments.Experiments show that the proposed system can easily get the images with high pixels,fast detection and high reliability.
出处
《计算机仿真》
CSCD
北大核心
2016年第5期161-165,234,共6页
Computer Simulation
关键词
纹理特征
智能监控
破损检测
脉冲耦合神经网络
交通线
Texture feature
Intelligent surveillance
Damage detection
Pulse coupled neural network
Traffic line
