Unmanned Vehicle Route Tracking Method Based on Video Image Processing.

Most unmanned vehicle path tracking methods ignore video image processing, resulting in a lot of interference information and severe distortion in the original video image, unable to accurately obtain road information, and reducing the accuracy of path tracking. This paper proposes a vehicle path tracking method based on video image processing. The original road condition image is filtered by median filtering method to reduce the interference of noise on image quality; the filtered road condition image is binarized to distinguish the image from the target image and the background image; and the boundary contour of the binary image is extracted by four neighborhood method to obtain the required road condition feature information. At the same time, the computational complexity is reduced. Based on the road condition characteristic information, the preview deviation angle and path curvature are calculated by preview point sequence; the driving speed is determined according to the path curvature, and the longitudinal control is realized; the preview deviation angle is converted into the control quantity of front wheel rotation angle by Pure Pursuit algorithm, and the lateral control is realized. The experimental results show that the driverless vehicle can track the reference trajectory quickly with different reference speeds, and then its position deviation is controlled within 0.05 m. The average energy consumption of path tracking is 355.13 J, which shows that the driverless vehicle using this method can achieve precise path tracking with low energy consumption. [ABSTRACT FROM AUTHOR]