| 凤鹏飞,金会庆,蒋玉亭.一种基于形态学与边缘点投票统计的车道线快速检测算法[J].唐山学院学报,2017,30(6):1-7 |
| 一种基于形态学与边缘点投票统计的车道线快速检测算法 |
| A Fast Lane Detection Algorithm Based on Morphology and Edge Point Voting Statistics |
| |
| DOI：10.16160/j.cnki.tsxyxb.2017.06.001 |
| 中文关键词: 形态学 边缘点投票统计 车道线检测算法 |
| 英文关键词: morphology edge points voting statistics lane detection algorithm |
| 基金项目:国家自然科学基金项目（51375131，51675151）；安徽省教育厅自然科学基金项目（KJ2016A890） |
|
| 摘要点击次数: 8101 |
| 全文下载次数: 6423 |
| 中文摘要: |
| 提出了一种基于形态学与边缘点投票统计的车道线快速检测算法，在道路图像感兴趣区域内进行数学形态学颗粒分析和骨架化，获取车道中心线，再进行车道边缘点筛选与投票，通过统计搜索的方式检测出车道线。实验采用数字信号处理芯片DSP为图像处理硬件开发平台，在软件系统CCS下调试程序。实验结果表明，该算法在车道偏离预警系统中运行具有较好的车道线检测效果，在复杂行驶环境下能正常运行，鲁棒性能较好。 |
| 英文摘要: |
| In this paper, a fast lane detection algorithm based on morphology and edge point voting statistics is presented, which can detect lane lines by conducting a mathematical morphological particle analysis of and skeletonizing the areas of interest of road images, obtaining the lane center line,and screening and voting on the lane edges, In the experiment, digital signal processing chip DSP is used as the hardware of development platform of the image processing, and the system is debugged with the software of CCS. The experimental results show that the algorithm can effectively detect lane lines in the lane departure warning system and the robust performance is good, even in the complex driving environment. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |