计算机与现代化

• 算法设计与分析 • 上一篇    下一篇

基于ORB和KLT的移动增强现实三维注册算法

  

  1. 1.厦门大学信息科学与技术学院,福建厦门361005;2.厦门大学软件学院,福建厦门361005
  • 收稿日期:2013-12-11 出版日期:2014-03-24 发布日期:2014-03-31
  • 作者简介:肖斌 (1990-),男,山东临沂人,厦门大学信息科学与技术学院硕士研究生,研究方向:增强现实,移动多媒体; 通信作者:江弋(1960-),男,副教授,研究方向:图像视觉; 林凡(1978-),男,厦门大学软件学院助教,研究方向:移动云计算,嵌入式。
  • 基金资助:
    国家自然科学基金资助项目(61001143)

Mobile Augmented Reality 3D Registration Algorithm Based on ORB and KLT

  1. 1. School of Information Science and Technology, Xiamen University, Xiamen 361005, China; 2. School of Software, Xiamen University, Xiamen 361005, China
  • Received:2013-12-11 Online:2014-03-24 Published:2014-03-31

摘要: 针对移动设备上现有注册算法速度较慢和实时性较弱的问题,提出一种基于ORB和KLT的移动增强现实三维注册算法,利用ORB进行特征提取和匹配,减少算法时间开销,同时使用改进的KLT算法对特征位置进行预测跟踪,改善移动三维注册算法中的实时性能。实验结果表明,在移动设备上采用ORB和KLT算法,不仅较好地解决移动增强现实三维注册实时性差的问题,同时也能保证有效的匹配性能。

关键词: 智能手机, 实时性, 特征跟踪

Abstract: 3D registration algorithm is the key technology of augmented reality, as well as the indicator of system performance. Since existing algorithm, such as improved SIFT operator and FASTSURF operator, is weak on realtime feature, this paper proposes an augmented reality 3D registration algorithm based on ORB and KLT, which reduces the time overhead and improves the matching performance. Experimental results show that the new algorithm which running on the mobile device, not only can reduce the time cost but also can ensure efficient matching performance.

Key words: smart mobilephone, realtime, feature tracking

中图分类号: