3D Reconstruction Based On Point Cloud And Edge Detection

With the development of computer graphics,digital image processing technology and scanning technology,three-dimensional reconstruction technology is also constantly improving.How to use three-dimensional reconstruction technology to reconstruct the surface of a damaged object and extract the inside area of a certain contour on the surface of the damaged object is of great significance to repair.For example,in the research of cultural relics protection,staffs need to use threedimensional reconstruction technology to obtain the current morphology and surface area of the damaged area of a cultural relic,then conduct related analysis and research on it,and finally formulate a restoration plan for the damaged cultural relic.Aiming at the extraction problem of the contour area on the three-dimensional surface,this paper proposes and realizes the contour area detection and extraction by converting the threedimensional contour detection into the detection of two-dimensional image,and then using the mapping relationship between the two-dimensional image and the triangle on the three-dimensional surface to extract the three-dimensional surface.First of all,this article improves the remeshing algorithm and write a parallel program by CUDA,which is about 10 times faster than using Open MP.As a result,it saves a lot of computing time.This article also implements a surface reconstruction algorithm based on the restricted Voronoi diagram;By using Open GL to render the 3D model as a 2D image,we retain the relationship between the 3D surface and the 2D image,and then we use the Canny algorithm to detect edges on the 2D image.The research finds that the use of remeshing can improve the effect of edge detection;We also proposed the mapping relationship with the two-dimensional image to extract the area within the contour of the three-dimensional surface.The experimental results show that a certain contour area of the three-dimensional surface can be successfully extracted by our method.Finally,the method in this paper is integrated into a system,and the software is written by Qt,which is convenient for users to use.