Full-Text Search:
Home|Journal Papers|About CNKI|User Service|FAQ|Contact Us|中文
《Optical Technique》 2018-05
Add to Favorite Get Latest Update

Camera relative orientation in Large-scale photogrammetry

WANG Jun;DONG Mingli;LI Wei;SUN Peng;Beijing Key Laboratory of Optoelectronic Test Technology,Beijing Information Science &Technology University;Institute of Information Photonics and Optical Communications,Beijing University of Posts and Telecommunications;  
In order to improve the measurement accuracy and stability of large-scale photogrammetry systems,a new relative orientation method based on local iterative search algorithm is proposed.The method calculates the relative orientation parameters between images of an image pair using only the coordinates of correspondences.The initial estimate of essential matrix between the two images is acquired through the closed-loop polynomial algorithm.The essential matrix is iteratively optimized to obtain the orthogonal matrix and epipolar error equations.All the possible relative exterior orientation parameters between camera positions are calculated by essential matrix decomposition,and the correct solution is determined by Cheirality constraint.The orientation experiment of an expandable truss microwave antenna profile measurement(6 m×4 m×3 m)is carried out to verify the accuracy and reliability of the new method.Compared with the traditional orientation methods,the proposed local iterative search algorithm provides more precise and stable orientation results.Translation error is less than 5 mm,rotation angle error is less than 0.1°.The method eliminates the orientation dependence on reference objects,which will play a key role to improve photogrammetric system performance.
【Fund】: 国家自然科学基金项目(51475046);; 北京市教委科技计划重点项目(KZ201711232029)
【CateGory Index】: TP391.41
Download(CAJ format) Download(PDF format)
CAJViewer7.0 supports all the CNKI file formats; AdobeReader only supports the PDF format.
©2006 Tsinghua Tongfang Knowledge Network Technology Co., Ltd.(Beijing)(TTKN) All rights reserved