计算机与现代化 ›› 2012, Vol. 1 ›› Issue (11): 214-216.doi: 10.3969/j.issn.1006-2475.2012.11.052

• 应用与开发 • 上一篇    下一篇

姿态估计算法在古建筑室内环境中的应用研究

师小燕1,万韬阮2,汤 汶3,李诗宁1,成红艳1,景军锋1,朱耀麟1   

  1. 1.西安工程大学电子信息学院,陕西 西安 710048;2.布拉德福德大学计算机信息媒体学院,英国 布拉德福德 BD7 1DP;3.提赛德大学计算机学院,英国 米德尔斯堡 TS1 3BA
  • 收稿日期:2012-09-20 修回日期:1900-01-01 出版日期:2012-11-10 发布日期:2012-11-10

Application Research on Pose Estimation Algorithm in Ancient Building Indoor Environment

SHI Xiao-yan1, WAN Tao-ruan2, TANG Wen3, LI Shi-ning1, CHENG Hong-yan1, JING Jun-feng1, ZHU Yao-lin1   

  1. 1. College of Electronics and Information, Xi'an Polytechnic University, Xi'an 710048, China;2. School of Computing, Informatics and Media, University of Bradford, Bradford BD7 1DP, UK;3. School of Computing, Teesside University, Middlesbrough TS1 3BA, UK
  • Received:2012-09-20 Revised:1900-01-01 Online:2012-11-10 Published:2012-11-10

摘要: 摄像机姿态估计是影响三维注册成功与否的关键技术,目前在基于基本矩阵的估计中最流行的是八点算法,但八点算法针对平面场景出现明显的退化,因此在针对古建筑室内环境的注册应用中,为了扩大场景的适用范围,利用计算机视觉技术,对基于五点算法和八点算法的两种算法进行研究。以盒子结构为古建筑的室内环境结构进行模拟,结果表明,五点算法较八点算法恢复的三维模型有更高的精确性。

关键词: 注册, 五点算法, 八点算法, 古建筑, 计算机视觉, 三维模型

Abstract: Camera pose estimation is the key technology of 3D registration. At present, based on the fundamental matrix estimation, the eight-point algorithm is the most popular, but the eight-point algorithm is obvious degradation in the case of planar scene. In order to expand the application scope of the scene in the ancient ruins of indoor environment registration application, the two algorithms based on the five-point algorithm and eight-point algorithm are studied by computer vision. Using the box structure as the environmental structure of ancient buildings, the experimental results show that five-point algorithm than eight-point algorithm has higher accuracy in the recovery of 3D model.

Key words: registration, five-point algorithm, eight-point algorithm, ancient buildings, computer vision, 3D model

中图分类号: