摘要
对于道路车道的识别提出了一种新的算法,主要包含两个主要部分:图像预处理部分首先通过图像灰度化和直方图均衡化得到灰度图像;然后进行二值化得到车道嫌疑区;最后通过形态学处理和边缘检测得到车道嫌疑区的边缘。车道检测部分通过对边缘图进行霍夫变化后利用车道线的位置特殊性来分别识别左右车道线。最后实验对比表明此算法能较大地提高准确性和实时性。
A novel method for lane marking recognition that includes two sections was introduced .The im‐age pre‐processing section got the gray image by using methods of gradation and histogram equalization .Bi‐narization was used to get the region of interest (ROI) .The boundary of ROI was obtained through mor‐phology operation and edge detection .The left and right lanes were detected in the lane detection section through Hough transform basing on the special positions of lanes .Experiment results indicate that this no‐vel algorithm features fast ,precise and reliable .
出处
《西安航空学院学报》
2015年第3期75-78,共4页
Journal of Xi’an Aeronautical Institute
关键词
车道识别
霍夫变换
图像处理
lane detection
hough transform
image processing
