摘要
提出基于关键帧匹配的增强现实跟踪注册算法,实现圆明园大水法遗址的数字重现。并采用随机树的方法将特征点的匹配问题转化为特征模式分类问题,解决图像间的宽基线特征匹配。基于该算法构建出基于视频透视式头盔显示器的移动增强现实系统。实验验证表明,该方法具有实时、鲁棒,适用于户外跟踪注册。
A keyframe based augmented reality registration algorithm was proposed applying for the digital reconstruction of the Yuanmingyuan heritage site. Randomized tree method was used to solve the wide-baseline matching of keypoints extracted from the input image to those keyframes as a classification problem. A video see-through mobile augmented reality system was built to test this algorithm. Experimental test demonstrates that this algorithm is real-time, robustness and effective for outdoor tracking.
出处
《系统仿真学报》
CAS
CSCD
北大核心
2010年第2期424-428,共5页
Journal of System Simulation
基金
国家高技术发展计划(863计划)(2007AA01Z325)
国家自然科学基金(60903070
60673198
60827003)
教育部长江学者和创新团队发展计划项目(IRT0606)
关键词
户外增强现实
跟踪定位
随机树
特征匹配
outdoor augmented reality
camera tracking
randomized tree
feature matching
