基于全局阈值分割及多项式拟合的车道线识别

Lane Line Recognition Based on Global Threshold Segmentation and Polynomial Fitting

为了在汽车辅助驾驶中准确得到车道线,本设计在阈值分割中比较了全局阈值、最大类间方差阈值两种方法,确定了基于全局阈值的方法来进行阈值分割。得到分割图后,进行透视变换,通过将像素按列相加的直方图得到左右车道线的基点,最后通过滑动窗多项式拟合左右车道线。仿真结果表明,该方法能够准确识别车道线,算法简单,运算效率高。

In order to accurately obtain lane lines in vehicle assisted driving,this design compares the global threshold and the maximum between-class variance threshold in the threshold segmentation,and determines the method based on the global threshold to perform threshold segmentation.After the segmentation map is obtained,the perspective transformation is performed,and the base points of the left and right lane lines are obtained through the histogram of adding the pixels in columns,and finally the left and right lane lines are fitted by the sliding window polynomial.The simulation results show that the method can accurately identify lane lines,with simple algorithm and high computational efficiency.