摘要
车道偏移的检测是智能车辆辅助驾驶系统中的重要技术问题之一。通过基于灰度阈值分割的梯度边缘检测技术,在对路面图像进行边缘检测的同时,配合以路面的灰度信息,准确地分离出车道标志线的边缘,再依此定义车道的跟踪区域——感兴趣区域(ROI),利用车道边缘信息定义边缘分布函数EDF(Edge Distribution Function),通过对跟踪区域中车道线梯度方向的分析,获取两条车道标志线在道路图像中的方向,以此作为车道偏移判断与预警的主要根据。该方法能够有效地抑制图像中非线性物体的干扰,是一种有效、可靠的车道偏移检测与预警方法。
The detection of lane departure is one of the most significant techniques in Intelligent Vehicle Navigation. With the grey information on the road, we accurately identify the edges of lane mark when detecting the edges of road image by the technique of grey-threshold segmentation based gradient edge detection from the road image. Then the precisely extracted lane mark will be utilized for the definition of Region of Interest (ROI) ,which is the area for lane tracking. The Edge Distribution Function (EDF) ,which is defined according to the edge information of lane mark,will be applied in the ROI to extract the lane mark's gradient information and consequently get the angles of two lane marks in the image. The final judgement and warning of the lane departure will mainly depend on the angles of the lane mark computed above. This processing can be efficient and reliable as a method for lane detection and tracking on account of its accuracy, stability and ability to prevent the disturbance from the non-linear object on the road.
出处
《计算机应用与软件》
CSCD
2009年第1期118-119,128,共3页
Computer Applications and Software
基金
江苏省高技术研究(G2005028)。