Vehicle motion track estimation method

Abstract

The invention relates to a vehicle motion track estimation method which includes the steps: S1 acquiring a body all-round looking image of a vehicle; S2 detecting angular points in the body image; S3tracking the angular points by an LK optical flow tracking method, and acquiring tracking points of the angular points in an image of next frame; S4 acquiring first motion information of the vehicle through a vehicular sensor and a vehicle motion model; S5 selecting the angular points based on the first motion information; S6 secondarily selecting the angular points to obtain an optimal matrix model, and calculating second motion information of the vehicle through the matrix model; S7 performing Kalman filtering fusion on the first motion information and the second motion information to obtaina vehicle motion track. By the vehicle motion track estimation method, high-accuracy vehicle motion tracks can be estimated when the vehicle runs at a high speed and a low speed.

Description

technical field [0001] The invention relates to the field of vehicle motion track estimation, in particular to a method for estimating vehicle motion track. Background technique [0002] At present, the estimation of vehicle motion trajectory includes two methods: image optical flow method and vehicle sensor. The vehicle trajectory estimation method based on the image optical flow method can accurately estimate the trajectory of the vehicle within a certain range of accuracy and stability requirements by estimating the image motion and the mapping relationship between the image and the actual world coordinates. When the speed of the vehicle is high, the accuracy of the vehicle trajectory will be estimated by using the image optical flow method, and the error of the estimated trajectory will be large. [0003] The method of obtaining information such as vehicle wheel speed and steering wheel angle based on on-board sensors, and then estimating the vehicle trajectory accordin...

AI Technical Summary

Problems solved by technology

[0004] (1) The accuracy of the estimated vehicle trajectory cannot be guaranteed. If the accuracy of the on-board sensor is low or the sensor is aging, the estimated accuracy of the vehicle trajectory will be low;

[0005] (2) There is a certain degree of noise in the trajectory. Due to the characteristics of the vehicle sensor itself and environmental factors, the data collected by the sensor is not stable and reliable data, and there is usually random noise interference, which affects the estimation result;

[0006] (3) When the speed of the vehicle is less than 5km / h, the applicable conditions of the vehicle motion model will no longer be met, therefore, the trajectory of the vehicle estimated at low speed is inaccurate;

[0007] (4) When the vehicle is in a large turn, due to the cumulative error of the sensor, the trajectory estimated by the vehicle model will deviate greatly from the actual position

Images