摘要
为了提高客运车辆危险驾驶状态的辨识效能,提出利用机器视觉辨识危险驾驶行为。使用车载CCD实时采集路面图像,依据图像处理算法检测车道标识线的位置以及本车与前方目标车辆的实时距离,建立了车辆横向偏航辨识模型和车辆纵向危险行驶状态辨识模型;结合所建模型的辨识结果,确定预警方案。实车试验结果表明,所建模型能有效辨识车辆横向偏航和纵向跟车过近危险行驶行为,可用来降低潜在的危险驾驶行为,提高车辆在途行驶安全性。
In order to improve the identification efficiency of dangerous driving status for passenger vehicle,an algorithm to identify dangerous driving behavior by using machine vision was proposed.Vehicle-mounted CCD was used to collect the real time digital road image.Based on the lane line detection,real time distance between host vehicle and preceding vehicle with the image processing algorithm,the vehicle lateral lane departure identification model and the longitudinal dangerous driving status identification model were built.Combined with the identification result of the related model,the warning method was selected.The experimental result shows that the model can effectively identify the lateral lane departure and the longitudinal dangerous driving behavior,and can reduce the potential dangerous driving behavior,improve the vehicle safety during the transportation.
出处
《汽车实用技术》
2015年第11期20-22,共3页
Automobile Applied Technology
基金
国家自然科学基金面上项目(51278062)
关键词
机器视觉
横向偏航
纵向危险行驶状态
辨识
Machine vision
lateral departure
longitudinal dangerous driving status
identification
