@inproceedings{4a2835bb423b4d848fabda74768d4c6f,
title = "A 3D reconstruction system for large scene based on RGB-D image",
abstract = "As an important research topic in the field of computer vision, 3D modeling of complex scene has an extensive application prospect. RGB-D sensor has been widely used to obtain the depth information in recent years. However, the existing processing system is simply suitable for small-scale scene modeling. In order to develop a better algorithm for large-scale complex scene modeling, this paper builds a 3D scene reconstruction system based on RGB-D images, achieving a better performance in accuracy and real time. SIFT algorithm is first to extract key points to match the descriptors between consecutive frames. By converting into three-dimensional space through the intrinsic matrix, the effective pixel points in the images are then reintegrated to establish the spatial point clouds model which is finally optimized by RANSAC algorithm. The experiments are based on the public database and propose the solution of the problems in the system, which provides a platform for basic research work.",
keywords = "3D reconstruction system, RANSAC, RGB-D",
author = "Hongren Wang and Pengbo Wang and Xiaodi Wang and Tianchen Peng and Baochang Zhang",
note = "Publisher Copyright: {\textcopyright} Springer Nature Switzerland AG 2018.; 8th International Conference on Intelligence Science and Big Data Engineering, IScIDE 2018 ; Conference date: 18-08-2018 Through 19-08-2018",
year = "2018",
doi = "10.1007/978-3-030-02698-1\_45",
language = "英语",
isbn = "9783030026974",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "518--527",
editor = "Kai Yu and Yuxin Peng and Xingpeng Jiang and Jiwen Lu",
booktitle = "Intelligence Science and Big Data Engineering - 8th International Conference, IScIDE 2018, Revised Selected Papers",
address = "德国",
}