The Research On Reconstruction And Optimization Of The Triangle Mesh Model Based On Industrial CT Sectional Images

The common acquisition methods of the three-dimensional (3D) data of productswere diverse in the fields of reverse engineering, such as coordinate measuringmachines (CMM),3D laser scanners and so on, which were typically used to measurethe outer surface of the product, while they can not measure the inner lumen data.Industrial computed tomography (Industrial CT or ICT), which can get both the outersurface of the product and the inner lumen data as an advanced non-destructive andnon-contact measure method. The measured data was a series of tomographic images ineither direction when using CT scanning the product, before using them for reverseengineer or innovative designs, you must obtain a three-dimensional model through thereconstruction of three-dimensional model. However, due to the large amount of dataobtained in industrial CT scanning, making a larger data redundancy afterreconstructing mesh. Because of the noise which produced during CT scanning and thedefects of reconstruction algorithm, making the reconstruction mesh exist noise anduneven surface. In order to solve the above problems, this paper used the scanned dataof CT to investigate the grid simplifying technology and the mesh smoothingtechnology, the main research contents were as follows:①In order to obtain high-quality3D model data, the primary task was to use theseries of industrial CT images to reconstruct the three-dimensional model. In this study,after a full analysis of the various advantages and disadvantages of thethree-dimensional model reconstruction algorithm, finally choose the moving cubealgorithms to reconstruct3D model, which meant through the three-dimensional datafield to strut the cube field, which was used to determine the internal and externalcontours, and then reconstructed the three-dimensional surface model.②In order to keep the grid model features when doing mesh simplification, andthen optimizing the mesh quality, based on triangle folding method, this study get thefolded position though the vertex forecasts and the Laplace operator adjustments, whichensured the quality of the simplified mesh. Introduction of distance error and angle errorin the choice of simplified triangle, which effectively reduced the operation errors ofmesh simplification and maintain feature of grid model.③In order to get a smooth mesh model, based on the characteristics of the reconstructedtriangular mesh model which was reconstructed though the use of industrial CT images, this paper proposed a global optimization method based on the grid model of implicit surface reconstruction.Firstly, calculated the normal vector of vertex though the normal vector of the triangular facets, andthen according to the location and coordinates of the vertex, using the Fast Fourier Transform tocalculate the Fourier coefficients of the surface features model function. After, we can get the modelsurface characteristic function by the inverse Fourier transform. Finally, selected the appropriatethreshold to extract isosurface.Results of this study have been integrated into the reverse software systems which wasdeveloped by research group, and though experimental verification, these methods can get smoothed,streamlined, and high-quality mesh model to meet the subsequent processing requirements.