A Research Of Four-wheel Positioning Technology Based On Machine Vision

Reasonable four-wheel positioning parameters play an important role in ensuring the stability of vehicle control,driving safety,reducing the wear degree of tire and suspension and reducing fuel consumption.Therefore,Therefore,it is of great significance for the four wheel positioning.Traditional four-wheel aligner design complex and numerous system modules,measurement error is large,the equipment is not easy to maintain.3D four-wheel aligner based on machine vision technology,simple operation,fast measurement,high accuracy,has become the current mainstream of four-wheel positioning products on the market,but the four-wheel aligner which is based on machine vision is expensive.Based on the study of the current 3D four-wheel positioning algorithm,this paper put forward the improvement measures on the current use of binocular vision technology four-wheel positioning,using monocular vision technology acquisition target image of the car front and rear wheels respectively,the space vector method is used as the positioning parameter calculation model and completed the calculation of the corresponding parameters of the wheel by image processing technology.This paper first introduces the background of the research,expounds the advantages of four-wheel aligner based on machine vision technology and the practical significance of the improvement of the current 3D four-wheel aligner.Then introduces the theory and algorithm of wheel positioning based on machine vision,emphatically discusses the space vector method wheel positioning model.It mainly completed the following aspects of the work: 1.In the camera calibration study,Zhang Zhengyou camera calibration method for the theoretical basis,the use of collected calibration plate image,through the calculation of the coherence matrix,combined with OpenCV camera calibration,The internal parameters of the camera and the distortion parameters are obtained;2.Under the premise of current wheel positioning technology based on machine vision is studied,this paper puts forward the scheme of wheel alignment technology based on monocular vision,the scheme is divided into two parts: image acquisition and data processing,and describes the composition and function of each part;3.Introduces the positioning parameters used in the calculation process of image binarization method,image gaussian smoothing filter method,image segmentation method and the center is used to extract the generalized Hough transform algorithm,and the related experiments of each algorithm perform;4.For the pose computational problems of the target board,using the matrix singular value decomposition and the method of least squares method combined with completed the posture parameters are calculated;5.Using the space vector method as thecalculation model to complete the calculation of the corresponding parameters of the wheel.Finally,the wheel positioning parameters are calculated by the method proposed in this paper,and then the results are compared with the measurement results of the same vehicle with a four-wheel aligner,which proves the validity of the method proposed in this paper,Which laid the foundation for the further study of four-wheel positioning technology based on monocular vision.