基于光流法的快速车道线识别算法研究

Algorithm Research of Fast Lane Detection Based on Optical Flow

摘要:

为提取无人驾驶车前方车道线信息，提出一种使用光流法的快速车道线识别算法。首先，根据连续视频帧之间的时间相关性，运用光流法检测车辆前方背景的相对移动。然后，利用车辆背景中特征点的移动方向和距离，对本帧图像中车道线的位置进行粗略定位，从而缩小本帧图像中车道线的检测区域，加速车道线识别算法。最后，通过对车道线像素点的处理，给出车道线类型信息。该算法提升了车道线检测算法的效率，降低了复合算子车道线检测算法的时间复杂度。在720*480像素下，算法实现了13.5Hz的处理速度，相较仅使用复合算子的处理算法提升了39.6%的处理速度，且算法检测效果良好。实车实验证明了算法的有效性和实时性。

Abstract:

In order to extract the lane line information of the driverless car, a fast lane line recognition algorithm using the optical flow method is proposed. First, using the temporal correlation between successive video frames, the optical flow method is used to detect the relative movement of the background in front of the vehicle. Then, using the moving direction and distance of the feature points in the background of the vehicle, the position of the lane line in the frame image is roughly positioned, thereby reducing the detection area of the lane line in the frame image and accelerating the lane line recognition algorithm. Finally, the lane line type is analyzed by processing the suspected lane line pixels. The improved algorithm improves the efficiency of the lane detection algorithm and solves the problem of high time complexity of the original multi-operator lane detection algorithm. At 720*480 pixels, the processing speed of 13.5 Hz is realized, which is 39.6% faster than the processing algorithm using only the composite operator, and the algorithm detection effect is good. The real vehicle experiment proves the effectiveness and real-time of the algorithm.