基于Canny算子和Hough变换的夜间车道线检测  

Nighttime Lane Markings Detection Based on Canny Operator and Hough Transform

作　　者：李亚娣[1] 黄海波[1] 李相鹏[1] 陈立国[1] LI Ya-di HUANG Hai-bo LI Xiang-peng CHEN Li-guo(College of Mechanical and Electrical Engineering, Robotics and Microsystems Center Soochow University, Suzhou 215000, P. R. China)

机构地区：[1]苏州大学机电工程学院机器人与微系统中心苏州纳米科技协同创新中心,苏州215000

出　　处：《科学技术与工程》2016年第31期234-237,242,共5页Science Technology and Engineering

基　　金：国家自然科学青年基金(61503270);国家自然科学青年基金项目(61203208)资助;国家自然科学基金(61327802)

摘　　要：针对夜间只有车灯照射路面图像整体较暗、光照不均匀、车道线不易检测的问题,提出了一种夜间车道线识别方法。首先,对预处理后的图像采用Laplacian算子进行边缘增强;然后,结合Otsu算法进行Canny边缘检测,再在边缘图像底部1/3区域中利用Hough变换进行直线拟合;最后,在斜率约束的基础上提出了一种内侧车道线提取算法,实现了车辆所在车道内侧车道线的检测。针对多种夜间车道线图像进行实验,结果表明,该算法准确提取出了内侧车道线。提出的方法能克服图像较暗和光照不均的影响,排除旁侧车道线、护栏等的干扰,有助于夜间车辆各行其道。It is difficult to detect the nighttime lane lines which showed darker and uneven lighted, for overco- ming these problems, a lane recognition method was proposed. Firstly, edge enhancement based on Laplacian was used to enhance the pre-processing images＇ edge. Then, image edge was detected by canny based on Otsu algo- rithm and the straight line of a third at the bottom of edge image was detected by Hough transform. Finally, an in- side lane line extraction algorithm was proposed on the basis of slope constraint and realized marking the inside lane that the vehicle driving on. By experimenting with various lane markers, this algorithm can realize the lane line de- tection. The proposed approach can overcome the influence of uneven light and be able to eliminate the interference from the side lane line, guard rail, etc. Lane line detection is conducive to the vehicle running on its road.

关 键 词：大津法 霍夫变换 边缘检测 车道线检测 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]