应用鲁棒卡尔曼滤波的车道线跟踪方法OA

To avoid detection errors caused by factors such as obstruction and wear of lane line parameters,this paper proposes a lane line tracking method using robust Kalman filtering.This method is based on the extracted discrete point data of lane markings and uses the cyclotron curve equation to construct a lane marking parameter model.Derive the observation equation of Kalman filter from uniformly distributed scatter data obtained through sampling preprocessing.Simultaneously combining the motion state data of the vehicle,derive the prediction equation of Kalman filter.And based on the residuals of the observation equation,adaptively calculate the covariance matrix of the system process noise.The experimental results show that the average lateral error of the left and right lane lines calculated by the method in this article is less than 9cm,and the maximum value does not exceed 25cm,which meets the functional requirements of intelligent driving scenarios.