IJMLC 2016 Vol.6(3): 160-166 ISSN: 2010-3700
DOI: 10.18178/ijmlc.2016.6.3.592

Global Hybrid Registration for 3D Constructed Surfaces Using Ray-Casting and Improved Self Adaptive Differential Evolution Algorithm

Linh Tao, Tinh Nguyen, and Hiroshi Hasegawa

Abstract—As a fundamental task in computer vision, registration has been a solution for many application such as: world modeling, part inspection and manufacturing, object recognition, pose estimation, robotic navigation, and reverse engineering. Given two images and set ones as the model, the aim is to find the best possible spatial transformation matrix causing 3D reconstruction of original object. The paper presents a new hybrid algorithm which improves both speed and convergence guarantee in comparison recently proposed methods of registering structured pointcloud surfaces by using a fast error calculation ray-casting based closest point method integrated with a new developed global optimization method Improve Self Adaptive Differential Evolution (ISADE). Ray-casting based L2 error calculation method enables the algorithm to find the local minima error while ISADE exploit the searching boundary to find the global minima. The new algorithm is evaluated to show the significant improvement in quality and robustness to state-of-the-art methods.

Index Terms—3D registration, ISADE, hybrid global registration, Ray-casting.

Linh Tao is with Shibaura Institute of Technology, Hasegawa laboratory, Japan (e-mail: taongoclinh@gmail.com).

[PDF]

Cite: Linh Tao, Tinh Nguyen, and Hiroshi Hasegawa, "Global Hybrid Registration for 3D Constructed Surfaces Using Ray-Casting and Improved Self Adaptive Differential Evolution Algorithm," International Journal of Machine Learning and Computing vol. 6, no. 3, pp. 160-166, 2016.

General Information

  • ISSN: 2010-3700 (Online)
  • Abbreviated Title: Int. J. Mach. Learn. Comput.
  • Frequency: Bimonthly
  • DOI: 10.18178/IJMLC
  • Editor-in-Chief: Dr. Lin Huang
  • Executive Editor:  Ms. Cherry L. Chen
  • Abstracing/Indexing: Scopus (since 2017), EI (INSPEC, IET), Google Scholar, Crossref, ProQuest, Electronic Journals Library.
  • E-mail: ijmlc@ejournal.net