Research On Calibration Method Of Large Size Measurement System Based On Line-structured-light Scanning

In modern manufacturing,there is an increasing demand for size measurement of large-size parts.Traditional detection methods have some problems such as low measurement efficiency and low accuracy.However,line structured light threedimensional visual size measurement is widely used in online measurement of large-size workpieces due to its simple structure,high measurement accuracy and fast speed.Largesize measurement systems is based on line structured light scanning,which are generally composed of line structured light vision sensors,linear motion platforms and industrial robots.The calibration of the pose relationship between them is one of the main factors that will affect the measurement accuracy of the system.This paper mainly studies the global calibration method of the line structured light three-dimensional scanning system and the large-scale measurement system.The main research contents are as follows:1.Established a global measurement model of the robot large-scale measurement system based on line structured light scanning.The composition structure of the largesize measurement system is analyzed,and the measurement models of the line structured light vision sensor and the line structured light scanning system are respectively established,and combined with a six-degree-of-freedom industrial robot,a large-size global measurement model is established.2.The calibration method of the linear structured light scanning system based on the spherical target is studied.First,a calibration model of the line structured light scanning system is established.Then,using the standard sphere as the target,the installation deflection angle of the line structured light sensor is calculated,and the calibration is completed by combining the movement distance of the line structured light sensor.The target point cloud feature extraction method is studied,and the Euclidean distance between the measurement points is proposed as a sample,and a least square point cloud feature extraction algorithm based on distance analysis is proposed to obtain the circle center coordinates.Two standard ball spacing measurement experiments are designed to verify the calibration accuracy of the linear structured light scanning system.3.Research on the global calibration of large-scale measurement system based on single point and multiple attitude.First,a global calibration model of the large-scale measurement system is established.Then,aiming at the calibration method of the relationship between the linear structured light scanning system based on single point and multi pose and the robot end pose,that is,the relationship between hand and eye.An improved global calibration algorithm based on outlier elimination is proposed to filter and eliminate the data involved in the calculation to obtain a more accurate pose matrix.Experiments show that the improved calibration method can greatly improve the measurement accuracy.A large-size measurement system was built,and measurement accuracy verification experiments were carried out.The hardware platform is built with KEYENCE line structured light sensor,linear guide rail,and FANUC six-degree-of-freedom robot.The corresponding system measurement software was developed using Visual Studio C#.The distance between two spherical targets with a distance of 1022.350 mm has been repeatedly measured,and compared with the third-party measurement results.The experimental results show that the relative error of measurement is less than 1.7‰ in the measurement range of 1020 mm.The calibration method studied in this paper can be used to calibrate the large-scale measurement system of the line structured light robot.