摘要
为了获得更加理想的车道线识别效果,以保证车辆的行驶安全,提出一种基于超像素和各向异性扩散相融合的车道线识别算法。首先从车道线图像中提取车道线识别的感兴趣区域,然后采用分水岭超分割算法提取感兴趣区域内的超像素,并通过各向异性扩散模型进一步实现车道线准确分割,最后采用仿真实验测试其性能。实验结果证明,本文算法提高了车道线识别的准确性,加快了车道线识别的速度,可以满足车道线识别的实时性,具有很好的鲁棒性。
In order to obtain more ideal recognition effect of the lane and ensure the driving safety of vehicles,this paper proposes a lane line recognition algorithm based on super-pixel and anisotropic diffusion.Firstly,interest region of lane recognition is extracted from the lane lines image,and then the watershed segmentation algorithm is used to extract super pixels from interest region,and the anisotropic diffusion algorithm is used to further improve segmentation accurate of the lane line,finally the simulation experiment is used to test its performance.The experimental results prove that the proposed algorithm improves the accuracy of lane recognition,speed up the lane recognition speed,and can meet the real-time lane detection and has good robustness.
出处
《激光杂志》
北大核心
2015年第4期122-125,共4页
Laser Journal
基金
福建省教育厅科研课题资助项目(JB13672S)
夏门城市职业学院教育改革研究资助项目(JGYJ1501)
关键词
智能交通
车道线识别
超像素
各向异性扩散
intelligent traffic
lane line recognition
super-pixel
anisotropic diffusion
