Four-wheel alignment of passenger cars is an indispensable process before the car leaves the factory. In this paper, a four-wheel positioning measurement structure design based on binocular vision is proposed, in which the two-dimensional image of the wheel is captured by a binocular camera, and the three-dimensional reconstruction technology is used to achieve non-contact measurement of wheel positioning parameters by segmenting and fitting the generated point cloud data. The method eliminates the need to install a target on the wheel, shortens the measurement time, guarantees a measurement accuracy of 0.1°, further optimises the point cloud data processing process, and accurately calculates the wheel's front beam angle, camber angle, main pin inner camber and main pin rear camber using the fitting results.
Zhao et al. (Thu,) studied this question.