Computer Science and Information Engineering, World Congress on (2009)
Los Angeles, California USA
Mar. 31, 2009 to Apr. 2, 2009
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/CSIE.2009.144
A fully automatic 3D reconstruction method based on images without manual modeling is presented in this paper. The method extracts 3D information from 2D images directly. Firstly, the feature points must be extracted from the images and then we match these corresponding points. Secondly, to determine the coordinates of 3D points, we should calculate the fundamental matrix, along with the extrinsic and intrinsic parameters of camera, which is to get the essential matrix. Consequently, we can implement the 3D reconstruction of the real scene after establishing the 3D model by Delauny triangulation and texture mapping and rendering in the late stage. Experiments with real images show that the method can reduce the reconstruction time of large-scale virtual scene greatly. It is characterized as efficiency, high stability and robustness.
3D reconstruction, epipolar geometry, features matching, Delauny triangularization, fundamental matrix, random sampling consensus (RANSAC)
J. Ze-tao, W. Wen-huan, W. Min and Z. Bi-na, "A Fully Automatic 3D Reconstruction Method Based on Images," 2009 WRI World Congress on Computer Science and Information Engineering, CSIE(CSIE), Los Angeles, CA, 2009, pp. 327-331.