Robust Range Image Registration using 3D lines

In this paper, we present an efficient method for accurate and automatic registration of two geometrically complex 3D range scans by using 3D lines. First we detect edges from 2D reflectance images and collect 3D edge contours by only taking into account valid foreground points. Then an efficient split-and-merge line fitting algorithm was used to detect 3D lines. Thus, we get a set of 3D lines from each 3D range scan. To efficiently match two sets of 3D lines, we compute the angles and shortest distances of pairs of 3D lines in each set which are invariant under rigid transformation, and build a fast search codebook for the set to be matched. An efficient RANSAC algorithm with robust transformation estimation using two sets of corresponding 3D lines is utilized to recover the rigid transformation between two scans. Finally we present experimental results which demonstrate efficiency and accuracy of our proposed method.

YAO Jian;  RUGGERI Mauro;  TADDEI Pierluigi;  SEQUEIRA Vitor; 

2010-11-16

IEEE - Institute of Electrical and Electronics

JRC58932

https://publications.jrc.ec.europa.eu/repository/handle/JRC58932,