基于方形边框标识的增强现实三维注册方法
|
摘要
基于计算机视觉的注册算法是增强现实系统最为关键的组成部分。针对人工标识注册的局限性与自然特征注册的速度限制,通过引入黑色边框的方式,结合自然特征设计了一种新的标识物注册方法。方法利用视频帧中标识物的边缘特征完成标识的初定位与快速跟踪,在标识中间区域提取ORB特征点,并利用随机抽样一致性(RANSAC)算法筛选得到的匹配点对,求解确定摄像机的位姿,利用前向后向光流跟踪优化检测效率。实验结果表明:注册方法在标识物运动情况下可准确完成注册,连续帧注册速率可达100帧/s。
The registration algorithm based on computer vision is the most crucial part of augmented reality system. In order to solve the limitation of artificial identification registration and the speed limitation of natural features registration,a new registration method by introducing black border and comibed with natural features is designed. This method uses edge feature of the marker in the video frame to achieve the approximate positioning and fast tracking,extract the ORB feature points in the middle area of the marker,matching point pair screened by random sample consensus algorithm( RANSAC) algorithm is used and solve and determine the pose of camera,use forward-backward optical flow to track and optimize efficiency of detection. Experimental result shows that the registration method can register accurately in the case of movement,registration efficiency of follow-up frame is up to 100 frame/s.
The registration algorithm based on computer vision is the most crucial part of augmented reality system. In order to solve the limitation of artificial identification registration and the speed limitation of natural features registration,a new registration method by introducing black border and comibed with natural features is designed. This method uses edge feature of the marker in the video frame to achieve the approximate positioning and fast tracking,extract the ORB feature points in the middle area of the marker,matching point pair screened by random sample consensus algorithm( RANSAC) algorithm is used and solve and determine the pose of camera,use forward-backward optical flow to track and optimize efficiency of detection. Experimental result shows that the registration method can register accurately in the case of movement,registration efficiency of follow-up frame is up to 100 frame/s.
引文
[1]Azuma R T.A survey of augmented reality[J].Rresence Teleoperators and Virtual Environments,1997,6(4):355-385.
[2]赵沁平.虚拟现实综述[J].中国科学(F辑:信息科学),2009,39(1):2-46.
[3]Kato H,Billinghurst M.Marker tracking and HMD calibration for a video-based aug-mented reality conferencing system[C]∥The2nd IEEE and ACM International Workshop on Augmented Reality,IWAR’99,San Francisco,CA,USA:IEEE,1999.
[4]田娟,郑郁正.模板匹配技术在图像识别中的应用[J].传感器与微系统,2008,27(1):112-114.
[5]Rublee E,Rabaud V,Konolige K,et al.ORB:An efficient alternative to SIFT or SURF[C]∥Proceedings of IEEE International Conference on Computer Vision,Los Alamitos:IEEE Computer Society Press,2011:2564-2571.
[6]Lucas B D,Kanade T.An iterative image tegistration technique with an application to stereo vision[C]∥Proceedings of the 7th International Joint Conference on Artificial Intelligence(IJCAI),San Francisco,CA,USA:Morgan Kaufmann Publishers Inc,1981:674-679.
[7]Zhang Z Y,Deriche R,Faugeras O.A robust technique for matching two uncal-ibrated images through the recovery of the unknown epipolar geometry[J].Artificial Intelligence,1995,78(1/2):87-229.
[8]Gonzalez R C,Woods R E.Digital image processing[M].New Jersey:PrenticeHall,1992.
[9]吴春福,陆国栋,张树有.基于拓扑映射的多边形顶点凸凹判别算法[J].计算机辅助设计与图形学学报,2002(9):810-814.
[10]Fischler M,Bolles R.Random sample consensus:A paradigm for model fitting with applications to image analysis and automated cartography[J].Communications of the ACM,1981,24(6):381-395.
[2]赵沁平.虚拟现实综述[J].中国科学(F辑:信息科学),2009,39(1):2-46.
[3]Kato H,Billinghurst M.Marker tracking and HMD calibration for a video-based aug-mented reality conferencing system[C]∥The2nd IEEE and ACM International Workshop on Augmented Reality,IWAR’99,San Francisco,CA,USA:IEEE,1999.
[4]田娟,郑郁正.模板匹配技术在图像识别中的应用[J].传感器与微系统,2008,27(1):112-114.
[5]Rublee E,Rabaud V,Konolige K,et al.ORB:An efficient alternative to SIFT or SURF[C]∥Proceedings of IEEE International Conference on Computer Vision,Los Alamitos:IEEE Computer Society Press,2011:2564-2571.
[6]Lucas B D,Kanade T.An iterative image tegistration technique with an application to stereo vision[C]∥Proceedings of the 7th International Joint Conference on Artificial Intelligence(IJCAI),San Francisco,CA,USA:Morgan Kaufmann Publishers Inc,1981:674-679.
[7]Zhang Z Y,Deriche R,Faugeras O.A robust technique for matching two uncal-ibrated images through the recovery of the unknown epipolar geometry[J].Artificial Intelligence,1995,78(1/2):87-229.
[8]Gonzalez R C,Woods R E.Digital image processing[M].New Jersey:PrenticeHall,1992.
[9]吴春福,陆国栋,张树有.基于拓扑映射的多边形顶点凸凹判别算法[J].计算机辅助设计与图形学学报,2002(9):810-814.
[10]Fischler M,Bolles R.Random sample consensus:A paradigm for model fitting with applications to image analysis and automated cartography[J].Communications of the ACM,1981,24(6):381-395.