An Efficient Hybrid Optimization Strategy for Surface Reconstruction

An efficient surface reconstruction strategy is presented in this study, which able to approximate non-convex sets of target points (TPs). The approach split two phases: (a) the mapping phase, making use shape preserving method (SPM) get a proper parametrization each sub-domain composing TPs set; (b) fitting where patch fitted by means suitable non-uniform rational basis spline (NURBS) considering, as design variables, all parameters involved its definition. To purpose, problem formulated constrained non-linear programming (CNLPP) defined over domain having changing dimension, wherein both number and value variables are optimized. deal with CNLPP, optimization process steps. Firstly, special genetic algorithm (GA) optimizes two-level evolution (species individuals). Secondly, solution provided GA constitutes initial guess for deterministic optimization, aims at improving accuracy surfaces. effectiveness proposed methodology proven through some meaningful benchmarks taken from literature.