Research On Vehicle Chassis Coordinate Measurement Based On Binocular Vision

Since the reform and opening up for more than 40 years,people’s life has become increasingly rich.Cars are no longer scarce goods.They have already entered thousands of households and become a part of people’s life.The following related industries,car maintenance,maintenance,are also rising.As a part of automobile maintenance,automobile chassis,because of its mechanical structure,as well as the complexity of the circuit,its maintenance is difficult and important,and the premise of maintenance is accurate measurement.In this paper,through the method of binocular stereo vision,a target based binocular vision measurement algorithm is studied,and the corresponding measurement system is formed.This paper first analyzes the research and development status of automobile chassis measurement,then analyzes the different measurement methods,and then determines the binocular stereo vision measurement method.Next,this paper analyzes the camera,lens and other key components,and completes the design of the related structure.The measurement system is divided into two parts: camera calibration algorithm and measurement conversion algorithm.A new calibration algorithm of binocular camera parameters is proposed,which is based on Zhang Zhengyou algorithm,with the help of coordinate measuring machine and particle swarm optimization artificial intelligence algorithm(PSO algorithm),to optimize the camera parameters.The practice shows that the new algorithm has significantly higher accuracy than Zhang Zhengyou algorithm,and slightly higher accuracy than Faugeras algorithm,which verifies the correctness and effectiveness of the proposed calibration algorithm.Aiming at the module of measurement conversion algorithm,this paper adopts generalized inverse matrix algorithm and Newton gradient descent algorithm,and verifies the accuracy and reliability of the algorithm through corresponding computer simulation.At the same time,in order to simplify the operation,using pyqt5,the corresponding software interface is designed,and the use process is described in detail.Finally,the algorithm verification experiment is designed to verify the reliability and stability of the corresponding algorithm with the help of coordinate measuring machine.The practice shows that the measurement error of binocular camera is less than 1.5mm in the depth direction of 2.5m-4.5m,and the single point repeatability is within 0.4mm.It is proved that the algorithm can realize the accurate measurement of vehicle chassis coordinates,which verifies the effectiveness of the algorithm.