摘要
为了提高车道偏离预警系统的检测精度及其便利性,提出了一种在感应区域内建立搜索窗口的车道识别算法,并利用数字图像处理及分析技术对所获道路图像进行滤波、增强及边缘识别处理,然后通过直接线性变换法将所获取的车道边缘信号的像素坐标转换为道路二维物方坐标,最后根据所获车道线坐标及汽车预设坐标显现汽车行驶状态,继而联系汽车驾驶状况,达到对汽车行驶状态的监控和对车道偏离的预警功能,具有良好的发展前景。
A lane detection algorithm is brought forward, which set searching windows in region of interest, and using filtering ,enhancement and edge detection processing in digital image processing and analysis technics to deal with the road images, then convert the obtained pixel coordinate of lane edge into 2-D coordinate of the road by direct linear transformation, Finally according to coordinates of lane and vehicle by default to visualize driving status, then link with current operation condition to achieve the function of monitoring driving situation and lane departure warning.
出处
《汽车工程师》
2011年第8期27-30,46,共5页
Automotive Engineer
基金
武汉理工大学自主创新研究基金(2010-zy-qc-017)
关键词
车道偏离预警系统
机器视觉
直接线性变换法
二维重建
Lane departure warning system
Machine vision
Direct linear transformation
2D-reconstruction
