A numerical optimization approach to generate smoothing spherical splines

JOURNAL OF GEOMETRY AND PHYSICS

卷:111

Article

Machado, L.¹  Monteiro, M. Teresa T.² 

[1] Univ Tras Os Montes & Alto Douro UTAD, Dept Math, P-5001801 Vila Real, Portugal

[2] Univ Minho, ALGORITMI Res Ctr, Dept Prod & Syst, P-4710057 Braga, Portugal

关键词: Geodesics;    Geometric smoothing splines;    Least squares;    Nonlinear constrained optimization;   

DOI  :  10.1016/j.geomphys.2016.10.007

来源: Elsevier

【 摘 要 】

Approximating data in curved spaces is a common procedure that is extremely required by modern applications arising, for instance, in aerospace and robotics industries. Here, we are particularly interested in finding smoothing cubic splines that best fit given data in the Euclidean sphere. To achieve this aim, a least squares optimization problem based on the minimization of a certain cost functional is formulated. To solve the problem a numerical algorithm is implemented using several routines from MATLAB toolboxes. The proposed algorithm is shown to be easy to implement, very accurate and precise for spherical data chosen randomly. (C) 2016 Elsevier B.V. All rights reserved.

【 授权许可】

Free