Research on 3D Reconstruction Optimization Method for High-Speed Flying Object Based on Multi-Granularity Matching
Longtao Fan, Tingting Yang, Sen Zhang · 2018
A three-dimensional(3D) reconstruction method based on multi-granularity matching is proposed by borrowing the symmetry feature for the high-speed flying Object, which can obtain the image information in the flight process. Firstly, the two-dimensional contour information of the object is extracted by two high-resolution CCD camera orthogonal image system. The image segmentation technique is used to extract the two-dimensional contour information of the object; thenimproved shape from shading (SFS) algorithm to solve the relative height and surface normal vector of each point on the surface of the object; and finally the 3D morphology of the surface of the high-speed flight is reconstructed by using the image splicing optimization technique. And verifies the feasibility and effectiveness of the proposed method through the three-dimensional reconstruction of the solid model.