摘要
分析了智能车辆安全辅助驾驶系统中弯道分道线的检测提取方法,提出一种基于道路区域分割的弯道检测新算法,包含道路区域分割和弯道边界检测。在分割出道路区域和天空区域并划定弯道检测的感兴趣区域后,提取分道线候选点,并对候选点进行校正,最终拟合并重建出弯道分道线,且准确判断了车道线弯曲方向。实验证明,该算法的实时性和准确性均高于在整幅图像中提取车道线的传统方法。
This paper described methods of bent lane line detection and extraction of intelligent vehicle safety assistant driving system.It introduced a new bent lane detection algorithm based on road segmentation,which was divided into road segmentation and bent lane detection.After dividing the road region and the sky region and determining the ROI for bent lane detection,extracted the candidate points,then correcting them,and finally fitted and reconstructed the bent lane lines.In addition,it judged out the curve direction of the bent lanes accurately.Experiments show that both the real-time performance and accuracy are improved comparing with the conventional method of extracting lane lines in the whole road image.
出处
《计算机应用研究》
CSCD
北大核心
2012年第7期2787-2789,共3页
Application Research of Computers
基金
国家自然科学基金资助项目(61104165)
关键词
弯道检测
道路区域分割
感兴趣区域
弯曲方向
bent lane detection
road segmentation
region of interest（ROI）
curve direction
