Structural Approximation Algorithms for One-dimensional Analysis of 3D-images in the Problem of Metrological Control of Surfaces of Large Industrial Products

The paper proposes an original version of structural approximation algorithms for one-dimensional analysis of 3D images of cellular surfaces of industrial products formed by laser scanning systems. The one-dimensionality of the analysis is achieved by decomposing the 3D image into lines of points corresponding to different positions of the laser beam when scanning the surface. Compared with existing algorithms, proposed algorithms have a higher speed and simplicity of setting parameters, values of which are determined by characteristics of products contained in the design documentation.