摘要
针对智能驾驶系统中车道线检测易受车道标记清晰度、光照能见度及道路环境复杂度等影响而导致车道线检测率不高的问题,提出一种特征结合的多阈值过滤车道线检测算法。即对图像进行梯度阈值过滤,再与颜色信息阈值过滤后的图像相结合,最后用改进Hough变换检测车道线。实验结果表明,本算法在存在阴影遮挡、路面出现泛白等因素下仍可以准确提取道路线信息,检测率平均高达93.89%,基本满足要求。
In order to solve the problem that lane detection can be easy to be affected by lane marking clarity, illumination visibility and road environment complexity in intelligent driving system, a multi-threshold filtering lane line detection algorithm is proposed in this paper. The original image is filtered by gradient threshold, and then combined with the image filtered by color information threshold. Finally, the improved Hough transform is used for lane line detection. Most importantly, the experiments show that the algorithm can extract the road line information accurately in the presence of poor lane marking, low illumination visibility, and shadow occlusion.The detection rate is as high as 93.89%, which can meet the requirements.
作者
刘雁斌
LIU Yanbin(Ningde Shachengwan Cross-sea Expressway Co., Ltd., Ningde 352200, China)
出处
《交通科技》
2019年第4期117-121,共5页
Transportation Science & Technology
