Share Email Print
cover

Proceedings Paper

A new mosaic method for three-dimensional surface
Author(s): Yun Yuan; Yongjun Ding
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

Three-dimensional (3-D) data mosaic is a indispensable link in surface measurement and digital terrain map generation. With respect to the mosaic problem of the local unorganized cloud points with rude registration and mass mismatched points, a new mosaic method for 3-D surface based on RANSAC is proposed. Every circular of this method is processed sequentially by random sample with additional shape constraint, data normalization of cloud points, absolute orientation, data denormalization of cloud points, inlier number statistic, etc. After N random sample trials the largest consensus set is selected, and at last the model is re-estimated using all the points in the selected subset. The minimal subset is composed of three non-colinear points which form a triangle. The shape of triangle is considered in random sample selection in order to make the sample selection reasonable. A new coordinate system transformation algorithm presented in this paper is used to avoid the singularity. The whole rotation transformation between the two coordinate systems can be solved by twice rotations expressed by Euler angle vector, each rotation has explicit physical means. Both simulation and real data are used to prove the correctness and validity of this mosaic method. This method has better noise immunity due to its robust estimation property, and has high accuracy as the shape constraint is added to random sample and the data normalization added to the absolute orientation. This method is applicable for high precision measurement of three-dimensional surface and also for the 3-D terrain mosaic.

Paper Details

Date Published: 18 August 2011
PDF: 9 pages
Proc. SPIE 8194, International Symposium on Photoelectronic Detection and Imaging 2011: Advances in Imaging Detectors and Applications, 81942T (18 August 2011); doi: 10.1117/12.900864
Show Author Affiliations
Yun Yuan, National Univ. of Defense Technology (China)
Hunan Key Lab. of Videometrics and Vision Navigation (China)
Yongjun Ding, National Univ. of Defense Technology (China)
Hunan Key Lab. of Videometrics and Vision Navigation (China)


Published in SPIE Proceedings Vol. 8194:
International Symposium on Photoelectronic Detection and Imaging 2011: Advances in Imaging Detectors and Applications
Makoto Ikeda; Nanjian Wu; Guangjun Zhang; Kecong Ai, Editor(s)

© SPIE. Terms of Use
Back to Top