Monocular-vision 3D Coordinate Measurement Technology Based On Handheld Target

In recent years,the demand for 3D coordinate measurement has become more and more important in the fields of industry,automobile,aeronautics and astronautics,national defense and so on.In the traditional method,3D coordinate measuring machine is used to measure the 3D coordinates of the key points of the object.This method has the advantages of high precision,simple operation and so on.The method of vision measurement is widely used in the field of 3D measurement because of its wide measuring range and high flexibility.At present,the main methods to measure the 3D coordinates of the object are the binocular-vision system and the multi-vision laser tracking system.In this article,a monocular-vision 3D coordinate measurement technique based on handheld feature target is studied.The external parameters of the target plane are calibrated by using circle dots.And the external parameters of the target plane are solved simultaneously with the internal parameters of the camera to acquire the 3D coordinates of the object.Aimed at the monocular-vision 3D coordinate measurement technology based on handheld target,the main contents and achievements of this paper are as follows:(1)In order to extract the ellipse center formed by the projection of the circular,this paper analyzes the sub-pixel detection algorithm of ellipse edge: interpolation and Zernike moment method.In order to obtain the pixel coordinates of the ellipse center according to the sub-pixel edge points of the ellipse,this paper studies the detection algorithm of ellipse center: centroid method,algebraic fitting method and geometric fitting method based on the principle of least squares.The experimental result shows that the Zernike moment method is more suitable for the sub-pixel edge detection.This paper also analyzes the advantages and disadvantages of two kinds of ellipse fitting methods,and determines the geometric fitting method with higher precision.(2)A monocular-vision 3D coordinate measurement model based on handheldtarget is established.Studying on the camera internal parameters calibration process,and on the basis of Zhang zhengyou's two step calibration method,two improvements are put forwarded: modify the projection direction of the distortion model,and consider two tangential distortions.And also,in this paper,the calibration process of the external parameters of the target plane are studied,two methods are proposed to optimize the external parameters of the target plane: image plane rotation method and target plane rotation method.(3)The optimization experiments of the target are designed.The influences of the correlative factors of the target on the calibration accuracy of the external parameters are studied.In order to evaluate the calibration accuracy of the external parameters of the target,the accuracy of the world coordinate calculation of the verification points is used.Based on the experimental data,the relationship between the number of feature points,the arrangement mode,the diameter,the distance between the baseline and the precision of the world coordinate calculation of the verification points are analyzed to provide an important basis for the optimal design of the target.(4)An experiment is designed to verify the accuracy of the model.According to the experimental data,the optimized target is made.This paper regards the two end points of the gauge block as the target points to be tested,and proposes to use the gauge block size as the standard to evaluate the model.Processing the collected image to acquire the 3D coordinates of the gauge block's two end points in the camera coordinate system,and calculate the size of the two end points,compare with the real value,and the accuracy of the model can be obtained.At last,the paper analyzes the causes of the errors and puts forward the improvement measures.