基于ORB-FREAK算法的移动增强现实系统实现
首发时间:2015-05-26
摘要:针对移动设备上现有注册算法存在运算量大、实时性差的问题,提出了一种基ORB-FREAK算法的移动增强现实三维注册技术。为缩减算法时间开销,该技术采用ORB和FREAK进行特征点提取和描述,通过描述符汉明距离进行匹配,再对匹配结果进行优化处理,然后实时估算得出摄像头位姿矩阵,最终在Android系统上实现了增强现实应用。实验表明,该方法可以有效地运行在普通智能手机上,并且在光照改变、尺度改变、不同角度、标志部分遮挡的情况下鲁棒性较好。
关键词: 移动增强现实 ORB-FREAK 汉明距离 Android
For information in English, please click here
Implementation of mobile augmented reality system based on ORB-FREAK Algorithm
Abstract:At present,the existing algorithm registration for mobile equipment is weak on real-time feature,this paper proposes an Tracking registration for Mobile augmented reality based on ORB-FREAK algorithm.To Reduce the algorrithm time overhead,using ORB and FREAK as detecter and descriptor,Hamming-based Brute-force Matcher,then optimize the matches results, according to the PNP algorithm for realtime estimates camera pose matrix, finally, in the Android system fulfilled an augmented reality application.Experiments show that this method can effectively runs in general intelligent mobile phone, and in the light of different angles, sign changes, partial occlusions ,it has a good Robustness.
Keywords: Mobile augmented reality ORB-FREAK Hamming distance;Android
基金:
论文图表:
引用
No.4641859105881814****
同行评议
共计0人参与
勘误表
基于ORB-FREAK算法的移动增强现实系统实现
评论
全部评论0/1000