基于Hough变换的高速公路车道线检测算法设计

Design of Highway Lane Detection Algorithm Based on Hough Transform

针对高速公路车道线的特征,提出了一种车道线检测识别方法.首先对图像进行小波提升分解,提取低频分解系数以此来降低因复杂环境因素对图像的影响;然后采用Canny算子进行边缘检测,根据边缘特征信息自动确定阈值对图像进行二值化处理;最后根据车道线的角度特征采用Hough变换对车辆内侧车道线进行检测.实验结果表明,该算法能够准确检测出高速公路车道线,提高其识别率.

Aiming at the characteristics of highway lane line,a method of lane line detection and identification is proposed.Firstly,the image is decomposed by wavelet transform and the low frequency decomposition coefficient is extracted to reduce the influence of complex environmental factors on the image.Then the Canny operator is used for edge detection,and the threshold value is automatically determined according to the edge feature information.Finally,Hough transform is used to detect the inside lane of vehicles according to the lane line Angle features.Experimental results show that the algorithm can accurately detect the lane information and improve its recognition rate.