Research Of Wheel Alignment Technology Based On Computer Vision

Vehicle wheel alignment parameters often change due to speed and road surface conditions factors, if wheel alignment parameters are abnormal that will lead to tire wear, driving wandering, steering heavy, increasing fuel consumption and other issues, directly affecting the vehicles running economy, comfort and safety. Traditional wheel alignment technique due to its principle limitations and complex operation process can not achieve the rapid detection. Computer vision technology has achieved successful applications in several fields, using computer vision to detect vehicle wheel alignment parameters does not require a large number of sensors and the complex calibration, simple operation for fast, accurate measurement of vehicle four positional parameters laid a theoretical foundation, has broad application prospects.In this paper, based on computer vision wheel alignment parameter detection theory and method to research, the movement of the wheel looks as a motion of rigid body of three-dimensional space, using rigid body kinematics theory to get rotary matrix and the rotary axis vector and rotary angle, give the wheel alignment parameters of the mathematical model and solution method. Under the conditions of the fixed camera, shooting on the target pattern mounted on the vehicle wheel when it moves front and back, and then collected image data for image processing and analysis, using SUSAN corner detection algorithm extracting feature points of image coordinates of target pattern when it moves front and back, through the camera calibration obtained the camera's intrinsic and extrinsic parameters and the homography matrix H. Then according to the mapping relation of the image coordinates and the world coordinates of target pattern to determine the world coordinates of feature points ,obtained the actual track of target pattern, and then to determine the actual trajectory of the wheel. Take the data into the mathematical model to calculate the wheel alignment parameters of each wheel position.In this paper, through the target pattern rotating at an angle to simulate the movement of the wheels. Use the simulation experiment to realize the detection of wheel alignment parameters. Verify the effectiveness and feasibility of the algorithm. At the same time, analyze the causes that the test results cause error, and gives measures to reduce the error, to facilitate the further work.