Analysis And Optimization Of Port Crane Photogrammetry System Accuracy

With the development of the port transportation industry,the number of port hoisting machinery is increasing,it is significance to test the size and structure of port hoisting machinery.The research shows that photogrammetry technology has the characteristics of low cost,high efficiency,and non-contact,which is very suitable for completing the measurement tasks of port cranes on the front line of the port.However,due to the limitations of factors such as the large size of the port cranes,the complex working environment of the port,and the limited location of the station in the actual measurement process,the actual measurement accuracy often differs from the ideal measurement accuracy and cannot meet the requirements.For this situation,this paper analyzes the development status of industrial photogrammetry technology at home and abroad,and uses numerical simulation analysis to analyze the mechanism of photogrammetry technology error for port cranes,and then conducts experimental verification to obtain the general Conclusions to guide survey operations.Based on the numerical analysis of the error generation mechanism,optimization algorithm is proposed.The specific research content is as follows:(1)Analyze the widely used camera calibration methods,and get the calibration parameters of the camera in this paper.The calibration calculation results are mapped on the image plane,and the general law of the degree of deviation of the actual imaging point from the ideal imaging point is obtained.(2)Establish a mathematical model of photogrammetry based on collinear equations and a photogrammetric structure parameter model based on the intersection of two rays in space,and analyze factors such as measurement distance,photographic baseline length,and measurement angle.Because both models have their own limitations,the calculation results of two models need to be supplemented and verified with each other,and finally a general conclusion is drawn.In the analysis process,consider the coupling relationship between the measurement angle,measurement distance and baseline length,we consider the impact of the measurement target size on the field layout parameters,and consider the impact of the measurement camera resolution on the measurement accuracy.Due to the imaging problem inside a single pixel,errors at the sub-pixel level of the image point cannot be avoided.Based on the analysis results of the error mechanism and the variation trend of the measurement error with the focal length calibration result,a real-time calibration method for focal length is proposed.Finally,experiments are designed,and lenses with different focal lengths were used to verify the conclusions calculated by numerical simulation analysis under different measurement distances and baseline lengths.(3)A weighted prediction optimization method is proposed for a specific straight line characteristic structure.By introducing weights in the calculation process,the coordinates of the measurement points on the straight line are optimized.This method does not depend on the accuracy of the input data.On the basis of the direct measurement results,the measurement results are corrected by fitting a straight line and finding predictive correction points on the straight line.Experiments show that this method has better optimization effect than direct measurement results.