Research On Machine Vision Measurement Method Of Automobile Central Channel Assembly

Traditional automobile production methods can no longer meet the rapidly changing market demand,and intelligent manufacturing has become an inevitable trend in the development of the automobile industry.Machine vision technology realizes the inspection and measurement of objects by replacing human vision,and has the characteristics of automation,high precision and high efficiency.The combination of machine vision measurement technology and automobile manufacturing technology is beneficial to the development and application of automobile intelligent manufacturing technology.In this paper,the automobile central channel assembly is always the object,and the application research of robot vision measurement technology in automobile body inspection is carried out.The main research contents are as follows.As the beginning of the research,the structural characteristics of the automobile central channel assembly and the difficulties in machine vision measurement are analyzed.According to the measurement requirements,the software and hardware platform of the robotic line structured light measurement system are designed and built.The line structured light measurement sensor in the measurement system is studied,and the structural parameter optimization design is carried out to ensure that the accuracy of the measurement sensor meets the measurement accuracy requirements of the automobile central channel assembly.The machine vision measurement system is calibrated.The line structured light plane calibration model and the calibration experimental platform for the line structured light measurement sensor are established,and a light plane calibration experiment and results analysis are completed.Based on the light plane calibration model,motion parameter calibration and hand-eye calibration models are established.The corresponding calibration data are obtained through calibration experiments to obtain the best calibration parameters.Aiming at the problem of light reflection on the metallic surface,the gray distribution of the cross section of the light strip are analyzed.An algorithm for extracting the center of the light strip based on double Gaussian fitting is proposed.The image processing flow of the light strip is formulated.The experiment and analysis that compared with traditional algorithms are carried.The credibility and accuracy of the algorithm is verified by the energy-based confidence evaluation method and the straight line fitting-based accuracy evaluation method.Measurement experiments are carried out by using the calibrated measurement system.A series of samples of known dimensions are measured,and the error distribution rule of the sensor’s movement direction and vertical movement direction is analyzed.It is concluded that the measurement system can meet the measurement requirement of ±0.5mm in the field of view of 50mm×50mm.The quality of point clouds data obtained by the light strip extraction algorithm based on double Gaussian fitting and the traditional algorithm is compared,and the effectiveness of the algorithm is verified.Measurement experiments were performed on the automobile central channel assembly,and stitching and feature fitting processes are performed,and the causes of measurement errors are analyzed.Experiments prove that the robot line structured light measurement system can meet the measurement needs of the automobile central channel assembly.The automobile central channel assembly measurement system and related research work have certain theoretical value and reference significance for the application research of machine vision measurement technology in automobile manufacturing.