The Key Techniques Of3D Reconstruction Based On Discrete Images

3D model has more advantage than2D image in sense of reality, and widely used in archaeology, education, medical, military, virtual reality and so on. How to quickly and effectively recover the scene of3D model from2D image has been an important research topic in the field of photogrammetry and computer vision, it includes many advanced algorithms in the field of computer vision. This paper deeply studies the3d reconstruction based on image of the related theory and algorithm and proposed a method of3D resconstion schme on the basis of the classical algorithm.This article first introduces the model of the camera, how to very few theoretical basis of3D reconstruction. Then designed two sets of feasible solution based on two images of3D reconstruction and multi-image of3D reconstruction and the various steps of solution are analyzed in detail. For feature extraction and feature matching process, this paper first introduces the Moravec operator, Harris operator and SIFT operator and expounds the characteristics of the corresponding matching method. Then this thesis introduces the RANSAC algorithm to eliminate false matching. For camera calibration steps, this paper introduces the traditional calibration method and camera self-calibration method and complete the discrete image calibration process using the Matlab toolbox. Steps for building a point cloud model, this paper designed two point cloud image reconstruction plan point cloud and image reconstruction. For point cloud grids and texture mapping procedure, this paper puts forward a multi-view texture mapping process by the intrducton of the point cloud grids and surface reconstruction method. Finally, the the result of reconstruction has been provided, which verified the feasibility of the scheme.