Work

Optimization-Driven Reconstruction of 3D Space Curves from Two Views Using NURBS

Year: 2025

Type: article

Abstract: In the realm of 3D curve reconstruction, Non-Uniform Rational B-Splines (NURBSs) offer a versatile mathematical tool due to their ability to precisely represent complex geometries. However, achieving high fitting accuracy in stereo-based applications remains challenging, primarily due to the nonlinear nature of weight optimization. This study introduces an enhanced iterative strategy that leverages the geometric significance of NURBS weights to incrementally refine curve fitting. By formulating an inverse optimization problem guided by model deformation principles, the proposed method progressively adjusts weights to minimize reprojection error. Experimental evaluations confirm the method’s convergence and demonstrate its superiority in fitting accuracy when compared to conventional optimization techniques.

Source: Mathematics

DOI: 10.3390/math13142256

Authors: Musrrat Ali, Deepika Saini, Sanoj Kumar, Abdul Rahaman Wahab Sait

Institutions: King Faisal University, Graphic Era University, University of Petroleum and Energy Studies

Language: English

Topic: Advanced Numerical Analysis Techniques

Subfield: Computational Mechanics

Field: Engineering

Domain: Physical Sciences

Open Access status: gold

Estimated APC paid: $3,137

Funder: Deanship of Scientific Research, King Faisal University