| In recent years,with the rapid development of manufacturing industry in our country,there are higher and higher requirements on industrial measuring techniques.As an important kind of industrial measuring system,industrial photogrammetry has the advantage of being non-contact,efficient and with high accuracy,and has a wide application in many fields.The optimization design of Industrial Photogrammetric Network(IPN)is of great importance in improving the precision,efficiency and automaticity of industrial photogrammetry.In this paper,a study is launched on this topic mainly from the following three aspects: the selection of datum,the network optimization design of IPN and the optimization of weight matrix.The main contents and innovations of this paper are as follows:1.The conception of IPN is defined and the characteristics of IPN are summarized.The importance of IPN is introduced through comparison with Control Networks of Foundation Pile.The research background,purpose and significance of this paper is introduced and the research status of the optimization design of industrial control network,of the precision analysis on industrial photogrammetry and of the camera station optimization design is summarized.2.The accuracy and reliability indexes of IPN are defined and the coordinate precision estimation model of marking points is established.The quality indexes of IPN are proposed from the aspects of accuracy and reliability based on adjustment model.Subject distance and intersection quality are defined,the influence regularity of subject distance and intersection quality on IPN's precision is analyzed,and the coordinate precision estimation model is established based on regression analysis.The influence of station number and network configuration on IPN's reliability is analyzed.3.The model of rank-defect self-calibration bundle adjustment based on quasi-stable datum is established.Firstly,the adjustment datum of IPN is introduced from the aspects of scale datum and position & pose datum.Then,based on the theory of self-calibration bundle adjustment and rank-defect free network adjustment,the G matrix of quasi-stable datum is designed,and the normal equation of rank-defect self-calibration bundle adjustment is established.Finally,the affection of different fixed datum and quasi-stable datum on adjustment result are compared through experiment,and the results show that: quasi-stable datum is more reasonable to be used in adjustment.4.An algorithm is designed to realize the optimization design of the network.The principles to set the network of IPN are summarized,and the influence of camera distribution on measuring precision is analyzed.The visuality,boundary of the available area,subject distance and so on are taken as constraint conditions,the mean square error of object points is taken as the fitness function,the position and posture parameters of cameras are taken as the fitness variables,based on adaptive genetic algorithm,the IPN optimization algorithm is designed.Simulation experiments and real test experiments are carried out,the results show that the algorithm is able to design measurement schemes with less station number and higher precision,and to provide guidance for actual photogrammetry process.5.The unweighted adjustment model of IPN is built.First of all,photographic light offset is defined,and the formula to calculate the photographic light offset is established.Then,by analyzing the influence of subject distance and image coordinates on photographic light offset,the rationality and necessity of assigning different weights to different image points is discussed,and the weighting formula of IPN adjustment is proposed.Finally,repeated measure experiment and the coordinate transformation experiment have been carried out,and the results show that the precision of weighted adjustment model is slightly better than the precision of unweighted adjustment model. |
PDF Full Text Request