A new pose measurement method based on microlens arrays

被引:1
作者
Sun, Changku [1 ,2 ]
Zu, Yahui [1 ]
Wang, Peng [1 ]
机构
[1] Tianjin Univ, State Key Lab Precis Measuring Technol & Instrume, Weijin Rd, Tianjin 300072, Peoples R China
[2] Luoyang Inst Electroopt Equipment, Sci & Technol Electroopt Control Lab, Luoyang 471000, Peoples R China
来源
2019 INTERNATIONAL CONFERENCE ON OPTICAL INSTRUMENTS AND TECHNOLOGY: OPTOELECTRONIC MEASUREMENT TECHNOLOGY AND SYSTEMS | 2020年 / 11439卷
关键词
Microlens arrays; pose measurement; optical marker;
D O I
10.1117/12.2542185
中图分类号
TH7 [仪器、仪表];
学科分类号
0804 ; 080401 ; 081102 ;
摘要
In pose measurement where infrared marking point is adopted as feature identification, only point coordinate information can be provided. In order to improve the accuracy of pose measurement, it is necessary to increase the number of feature points involved in pose solving, which limits the application of pose measurement method. A new pose measurement method based on two-dimensional optical coding marking orientation is designed. In the process of measurement, not only the coordinate information of the marking point can be used, but also the orientation information of the visual sensor to shoot the marker can be obtained, which improves the accuracy of pose measurement. The optical marker in this method includes a microlens array, a back two-dimensional code pattern and a background light source. When the visual sensor shoots the optical marker from different angles, different two-dimensional coding patterns can be obtained, which provides orientation information for pose measurement. Firstly, the images captured from different viewpoints are preprocessed, including segmenting the optical marker with the surrounding environment and transforming the images into undistorted frontal views. Then, on the basis of the pixel information of the microlens array's surface pattern, every single microlens for different viewpoints is encoded to obtain a set of code strings varying with the viewpoint. Finally, the code string information of different viewpoints is processed and the orientation model of two-dimensional optical coding marker is established. In addition, the factors that may impact the pose measurement accuracy are evaluated.
引用
收藏
页数:10
相关论文
共 10 条
[1]   Bokode: Imperceptible Visual tags for Camera Based Interaction from a Distance [J].
Mohan, Ankit ;
Woo, Grace ;
Hiura, Shinsaku ;
Smithwick, Quinn ;
Raskar, Ramesh .
ACM TRANSACTIONS ON GRAPHICS, 2009, 28 (03)
[2]   Monocular head pose estimation using generalized adaptive view-based appearance model [J].
Morency, Louis-Philippe ;
Whitehill, Jacob ;
Movellan, Javier .
IMAGE AND VISION COMPUTING, 2010, 28 (05) :754-761
[3]   Pose Hashing with Microlens Arrays [J].
Schillebeeckx, Ian ;
Pless, Robert .
COMPUTER VISION - ECCV 2016, PT III, 2016, 9907 :600-614
[4]  
Sun Changku, 2015, PLOS ONE, V10
[5]  
Tanaka H., 2012, 2012 IEEE 1st Global Conference on Consumer Electronics (GCCE 2012), P217, DOI 10.1109/GCCE.2012.6379585
[6]  
Tanaka H, 2012, INT C PATT RECOG, P837
[7]  
Tanaka H, 2012, IEEE INT C INT ROBOT, P4192, DOI 10.1109/IROS.2012.6386052
[8]   Improvement of accuracy for 2D marker-based tracking using particle filter [J].
Uematsu, Yuko ;
Saito, Hideo .
17TH INTERNATIONAL CONFERENCE ON ARTIFICIAL REALITY AND TELEXISTENCE, ICAT 2007, PROCEEDINGS, 2007, :183-189
[9]   Study on the position and orientation measurement method with monocular vision system [J].
Wang, Peng ;
Xiao, Xu ;
Zhang, Zimiao ;
Sun, Changku .
CHINESE OPTICS LETTERS, 2010, 8 (01) :55-58
[10]   Visual marker detection and decoding in AR systems: A comparative study [J].
Zhang, X ;
Fronz, S ;
Navab, N .
INTERNATIONAL SYMPOSIUM ON MIXED AND AUGMENTED REALITY, PROCEEDINGS, 2002, :97-106