Wheel Alignment Vision Method Based On The Parallel Constraint

With the rapid development of national economy, especially the transportationbusiness, the ratio of the medium truck and heavy truck have a greatly rise in theproportion of the car holding volume. Vehicle periodically wheel alignment detectionis the important means to guarantee the vehicle’s performance, economical efficiencyand security. Now, the domestic cars’ wheel alignment technology is much moremature, but the multi-axle structures of the large vehicles and the huge car bodiesneed a specialized wheel alignment detection technology to measure.This article starts from national wheel alignment detection methods of the largevehicles, and the exist problems for the using and safeguarding aspects. It makes adetailed analysis to the current status of the wheel alignment detection technologies athome and abroad, the principles of the wheel alignment detection method and theeffects of each positioning parameters to the vehicles’ performance. It introduces theprinciple of the non-contact wheel alignment detection method using the mainstreamcomputer visual technology, and it put forward a new wheel alignment method directto the large vehicles with huge bodies or multi-axle structures based on that principle.The method can be used to the trailers for axle parallel detection, and also to otherlarge vehicles for toe-camber detection. The paper focused on the need for theimprovement of the accuracy and speed of the algorithm in the process of solving themonocular vision object position and orientation. The algorithm simplifies thecomputational model, reduces the number of iterations, and improves the speed in thesolving process. The experimental results show that it has a good precision andenhances the solution speed. It provides a new method for real-time on-linemeasurement.