Research And Development Of Vehicle Wheel Alignment Parameters Vision Detection System

Visual detection technology is rising rapidly in recent years. It is a non-contact, fast,high precision, advanced measurement technology. It is widely used in Online testing,automated production. Wheel alignment parameters test is an important part of vehicledetection and it is an important means of safe driving. As the most advanced technologiesin the field, V3D based on computer vision attracted wide attention. The purpose of thepaper is to establish more effective wheel alignment parameters detection model based onmonocular vision, and build a complete visual inspection system. To complete the targetimage acquisition, feature information extraction, position calculation, dynamically outputwheel status and location data in three dimensions of the whole process.Firstly, wheel alignment parameters, detection methods, the position detection theoryand methods based on computer vision have been studied. Models of wheel alignmentparameters based on monocular vision mathematical have been constructed. Thecalculation of relationship between the target position and wheel alignment parameters hasbeen determined and a specific model solution method has been given. Secondly, designingthe overall structure of the WADS wheel alignment parameters visual inspection system,and details system functions, hardware and software architecture, testing processes,focusing on initialization of image acquisition card, parameter setting, interface softwaredesign and features information extraction of target image. In addition, building a realistictesting environment supports dynamic output wheel status in three dimensions. The paperdetails the parametric modeling methods of the basic components of vehicle chassis basedon OpenGL. Put forward the surface modeling of the tire tread based on two step mapping.Finally, several models based on WADS platform have been detected, and theexperimental results have been analyzed. Test results are basically the same with thepositioning results of current mainstream products. This indicates that the detection modelestablished in this paper is correct and effective, and the WADS system is feasible.