Research On The Algorithm And 3D Reconstruction Of Nondestructive Measurement Based On Slice Volume

Non-destructive measurement is the basis of reverse engineering technology.By adopting multiple methods to carry out multi-faceted technical measurement,existing geometric products or design drawings can be directly converted into abstract digital or engineering models.To solve the problem that the point cloud data model is difficult to solve in the three-dimensional non-destructive measurement method based on slice volume,this paper uses the improved particle swarm and moving tetrahedron algorithm to complete the solution and reconstruction of the point cloud data,and provides a solution for the reverse development of geometric products.A feasible method.The main contents of this article are as follows:Firstly,construct a mathematical model for solving the volume of geometric product slices.Through comprehensive analysis of the measurement principle and measurement process based on the grid layer volume non-destructive measurement method,the solution functions in 9 measurement directions of the geometric product container are constructed respectively.Combined with the characteristics of the product slice volume,an optimized mathematical model of grid slice volume information composed of 9n+n~3equations is constructed,which provides a prerequisite for the optimal solution of the geometric product slice volume mathematical model.Secondly,Then,the optimal solution of the geometric product mathematical model.The principle of the basic particle swarm optimization algorithm is analyzed comprehensively,the disadvantages of premature algorithm solving,low convergence accuracy are pointed out,and the influence of each parameter selection of the algorithm on the performance of the algorithm and the improvement strategy are given.Combined with the characteristics of the mathematical model,the BPSO with different inertia weight improvement strategies was solved and analyzed,and an improved discrete particle swarm optimization algorithm based on mesh slices was designed.By changing the particle size and increasing the variable dimension,it is found that the improved discrete particle swarm optimization algorithm has a better effect on solving the model based on the grid slice volume nondestructive measurement,and has significantly improved the convergence of the fitness function and the calculation efficiency.The improved algorithm Has a good search performance.Through the solution of the 1 million variable mathematical model,it is found that the accuracy of the solution error is 6.2%,which meets the requirements of engineering applications.Thirdly,the fitting and reconstruction of the geometric product point cloud model.The K-D tree and grid method are used to perform noise reduction and reduction on the data points of the large-scale point cloud model to obtain pre-processed point cloud data.In view of the shortcomings of the ambiguity of the moving cube algorithm(MC)and the specific characteristics of the data points of the point cloud model of this subject,this paper uses the MT algorithm based on the improved MC algorithm to process the point cloud data,and finally completes the three-dimensional model of simple mechanical workpiece Reconstruction of the fit.Finally,a mathematical model solution system for non-destructive measurement of slice volume was developed and the feasibility of this method was verified with specific engineering examples.By improving the discrete particle swarm algorithm to analyze and process the measurement data of actual mechanical workpieces,a point cloud data model of geometric products is obtained.Using MT algorithm to achieve point cloud gridding,through smoothing the reconstruction model,to achieve three-dimensional reconstruction of the entity,and from the experimental results to verify the feasibility of the grid slice volume measurement method.