The Research Of Vision Measurement System Calibration Method With Large Field Of View

Because of its advantages of non-contact,high efficiency and global information acquisition,vision measurement technology has been widely applied to the measurement of geometric quantity,and high precision and high efficiency calibration is the basis of accurate and rapid measurement of vision measurement system.However,the existing calibration method of vision measurement system faces low calibration efficiency and is not suitable for large field calibration.At the same time,the large field calibration error is large and the environment adaptability is poor in the field environment.In addition,there are some problems in calibration system,such as low calibration accuracy,low calibration efficiency and low calibration robustness.With the rapid development of our country's vision measurement technology,the requirements for the calibration method of the large-field visual measurement system are more and more harsh.Therefore,how to realize high-efficiency and high-precision calibration of vision measurement system in a large field of view is an urgent problem to solve in the field of visual measurement.It is of great significance to develop a calibration method of a large-field visual measurement system with good environmental applicability.In this paper,aiming at the key technical problems that still affect the further development of large-field calibration method,this paper puts forward the calibration method of large-field visual measurement system,which covers the application of "ensuring calibration efficiency,ensuring calibration accuracy and improving calibration robustness.A wide range of wide-field visual measurement system parameters rapid calibration method for large-field visual field measurement system parameters high-precision calibration method,as well as high-precision large-field measurement system parameters online on-line calibration method,the specific research contents are as follows:(1)Aiming at the problem of low calibration efficiency and poor applicability of the existing calibration methods in large field of view,a fast calibration method of large field measurement system based on scene information constraints proposed.Based on the characteristic information of parallel lines existing in the scene of the environment,the correlation between the scene information such as parallel lines and the parameters of the measurement system revealed.The linear constraint between the parameters of the orthogonal blanking information and the measurement system established to realize the rapid calibration of the measurement system parameters the distortion correction index function of measurement system was constructed and a camera distortion calibration method based on collinear constraint was formed,which eliminated the influence of calibration image distortion on calibration accuracy.Experimental results show that the fast calibration method of large-field measurement system based on scene information constraints proposed in this paper can calibrate the measurement system parameters with only one image.Compared with the widely used Zhang calibration method,the calibration time reduced by 80%and the calibration efficiency obviously improved without having to rely on high-precision calibration material,which has strong system applicability.Verify the certain calibration accuracy,and calibration efficiency.(2)Aiming at the problem that the field environmental factors greatly affect the calibration accuracy of large field of view,a high precision calibration method of large field visual measurement system based on orthogonal motion compensation is proposed.The results show that the calibration error increases by 0.3%within the range of 1 ° orthogonal error.In this study,the orthogonal degree of feature points and the camera internal parameter calibration method based on the orthogonal degree compensation proposed.The camera internal parameter calibration method based on the arbitrary angle shift proposed.The high precision calibration method of the internal parameters of the measurement system is proposed.The model of centroid transformation is established and the parameter deviation function of measuring system is constructed.The global parameter calibration method of measuring system based on centroid transformation propose,which eliminates the influence of the calibration error of internal parameter on the final calibration accuracy and improves the calibration accuracy with robustness.The method in the 5m × 8m measurement field of view,the system calibration error is only 0.09%.Compared with the traditional calibration method without orthogonality compensation,the calibration error reduced by 61%and the environmental adaptability is strong.(3)Aiming at the problem that measurement conditions easily change due to measurement environment factors,a field calibration method for large field of vision system with multi-constraint information fusion in complex environment proposed.Aiming at the problem of poor accuracy retention of the calibrator in complex field environment,the mutual reference information index is constructed and the parameter index function is established.A fast calibration method of the measurement system parameter based on the mutual reference information proposed,the accuracy of the calibration results is also discussed.At the same time,aiming at the difficulty of matching the size of the calibration object and the calibration field in the large field of view,a global three-dimensional information fusion function is established and the measurement parameters of the whole system based on the global three-Calibration method.In this paper,the calibration error of field calibration of large-field-of-view vision system with multi-constraint information fusion under complex environment is only 0.09%,only one calibration object image is needed.Compared with the traditional Zhang's calibration method,the calibration error is reduced by 34%and the stability of the calibration result is improved by 83%,breaking the boundary between the field of view and the size of the calibration object and solving the problem that the calibration accuracy Declining problems,effectively improving the calibration efficiency.Finally,using the proposed method for calibration of large-field visual measurement system,calibration and calibration experiments carried out in different measurement environments to verify the environmental applicability of the calibration method for large-field visual measurement system proposed in this paper.The experimental results show that the calibration method proposed in this paper has a good environmental applicability and can calibrate and calibrate the measurement parameters in different environments with high accuracy and efficiency.