Multi-View Registration Based on Triangle Constraint and Global Optimization

Abstract:

The measurement technology of 3D scanning in automation level is low, its measurement process is not stable, and the measurement error is large. After the principle of binocular structured light measurement was analyzed, match the marked point to multi-view registration based on the relative distance of any two marked point in space is unchanged. A multi-view registration is proposed using triangle constraint and global optimization. The method includes: remove the false matching points in possible matching points by joining triple constraint; after the first registration, use the center of marking points to global optimization through least squares principle method. The realization of the algorithm by use of tools such as VC++ and Open CV demonstrates that this method has the advantage of better restraining the accumulative error, keeping the process stable, and fast and realizing automatic registration.

• Periodical: Applied Mechanics and Materials (Volume 722)
• Pages: 327-333
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.722.327
• Online since: December 2014
• Authors: Rong Hua Wang*, Hua Zhang, Feng Xu, Li Yue
• Keywords: Binocular Structured Light Measurement, Global Optimization, Multi-View Registration, Relative Distance Unchanged, Triangle Constraint

References

[1] SOON-YONG PARK, MURALI SUBBARARAO, A multiview 3D modeling system based on stereo vision techniques, J. Machine Vision and Applications, 2005, 16(3): 148-156.

DOI: 10.1007/s00138-004-0165-2

[2] SONG Zhang. Recent progresses on real-time 3D shape measurement using digital fringe projection techniques, J. Optics and Lasers in Engineering, 2009, 48(2): 149-158.

DOI: 10.1016/j.optlaseng.2009.03.008

[3] SHAO Wei, MENG Lin, QIN Haohua, etal. Error analysis of the motion estimation based on SVD theory, C. /Proceedings of the 33th Chinese Control Conference, 2014: 915-921.

[4] GHOUTI, L. Robust perceptual color image hashing using quaternion singular value decomposition, C. /2014 IEEE International Conference on Acoustics, Speech and Signal Process, 2014: 3794-3798.

DOI: 10.1109/icassp.2014.6854311

[5] MAHMOOD N, OMAR C, TARDI T. Multiviews reconstruction for prosthetic design, J. The International Arab Journal of Information Technology, 2012, 9(1): 49-54.

[6] ZHOU Langming, ZHEN Shunyi, HUANG Rongyong. A Registration Algorithm for Point Clouds Obtained by Scanning Objects on Turntable, J. Journal of Surveying and Mapping, 2013, 42(1): 73-79.

[7] HUANG Ge, LI Xiaofeng. Fully Automatic Point Cloud Registration Algorithm Based on CEGI and Fourier Transformation, J. Journal of Sichuan University(Engineering Sciences), 2014, 46(05): 104-109.

[8] CHEN Yong, LIU Wenshong, LIU Shanwei, LIU Jinping. An Image Mosaics Software Based on Harris and RANSAC Algorithm, J. Surveying and Spatial Geographic Information, 2014, 37 (09): 97-100.

[9] YANG Fan, QUAN Wei, BAI Baoxing, XUE Yaohong. Automatically stitching 3D point-cloud based on marked point, J. Computer Engineering and Design, 2014, 35(09): 3321-3325.

[10] ZHANG Guangjun. Machine Vision. Bejing: Science Press, (2005).

[11] JIE Zexiao, GAO Xiang, CUI Jian. Extraction Algorithm of Circular Targets Used for Mobile Three-Dimensional Measurement, J. China Laser, 2012, 40(12): 1-6.

DOI: 10.3788/cjl201340.1208002

[12] XU Xiaodong, ZHAO Yi, LI Congxin. The Technique about Registration of Multi-view Data in the Optical Digitizing Measuring and Algorithm Realization, J. Machine with Hydraulic, 2005(10): 137-141.

[13] ZHOU Bo, CHEN Xinhe, XIE Xuedong. Multi-view cloud data automatically registration algorithm, J. Automation and Instrumentation, 2007(1): 11-13.