A hybrid-optimization method for assessing the realizability of wireframe sketches

  • Authors:

  • Philip Azariadis;Sofia Kyratzi;Nickolas S. Sapidis

  • Affiliations:

  • Department of Product and Systems Design Engineering, University of the Aegean, Hermoupolis, Syros, Greece;Department of Product and Systems Design Engineering, University of the Aegean, Hermoupolis, Syros, Greece;Department of Mechanical Engineering, University of Western Macedonia, Kozani, Greece

  • Venue:
  • 3D Research
  • Year:
  • 2013

Quantified Score

Hi-index 0.00

Visualization

Abstract

This paper introduces an optimization strategy for evaluating the realizability of a 2D wireframe sketch that conveys geometric and topological information about a 3D solid model. Applying the cross-section realizability criterion, one is able to assert whether a wireframe sketch is a true orthogonal projection of a 3D-solid. In this work, we first review current sketch interpretation methods and realizability criteria, and then we focus on an algebraic system derived from the cross-section realizability criterion. A two-phase hybrid-optimization approach for deriving cross-sections of a given wireframe sketch is introduced. In the first phase, a Genetic Algorithm is employed to produce an initial solution (i.e., an initial cross-section), which is refined by a Conjugate Gradient method in the second phase of the proposed approach. The final cross-section is an accurate solution of the aforementioned algebraic system. Then we are able to test sketch's realizability utilizing four criteria which are derived from the cross-section realizability criterion and the applied optimization procedure. The proposed optimization strategy is tested on wireframe sketches with accurate geometry and also on wireframe sketches with inaccurate geometry. Experimental numerical results are presented to illustrate the effectiveness and robustness of the proposed method.