Automatic multi-view registration of unordered range scans without feature extraction.pdfVIP

Neurocomputing 171 (2016) 1444–1453 Contents lists available at ScienceDirect Neurocomputing journal homepage: /locate/neucom Automatic multi-view registration of unordered range scans without feature extraction Jihua Zhu, Li Zhu n, Zhongyu Li, Chen Li, Jingru Cui Xian Jiaotong University, Xian 710049, China article info Article history: Received 1 June 2015 Received in revised form 14 July 2015 Accepted 28 July 2015 Communicated by Shaoting Zhang Available online 6 August 2015 Keywords: Multi-view registration Genetic algorithm Overlapping percentage Range scan Pair-wise registration abstract For 3D object modeling, this paper proposes an automatic approach for multi-view registration of unordered range scans by both the coarse and ?ne registration. In the coarse step, a spanning tree can be constructed by applying the presented approach for pair-wise registration of partially overlapping scans with the genetic algorithm (GA). To construct the spanning tree, the dual criterion is proposed to judge the reliability of pair-wise registration results. In this case, the ?rst scan can be selected as the root node and other scans can be added by the breadth-?rst search with the reliable results of the pair-wise registration. Subsequently, coarse results for multi-view registration can be calculated from the constructed spanning tree. In the ?ne step, the coarse registration results can be viewed as the initial parameters of the trimmed iterative closest point (TrICP) algorithm to obtain the accurate object model. Without any feature extraction, this approach can automatically achieve the multi-view registration of unordered range scans. Experiments were performed on public datasets to show its superiority on robustness for multi-view registration of unordered range scans. 2015 Elsevier B.V. All rights reserved. 1. Introduction Three dimensional (3D) modeling for real world has numerous application areas including digital museums [1], reverse engineering [2], robot mappin