A Genetic Algorithm based Optimization Method in 3D Solid Reconstruction from 2D Multi-View Engineering Drawings

There are mainly two categories for a 3D reconstruction from 2D drawings: B-Rep and CSG that both these methods have serious weaknesses despite being useful. B-Rep method which has been older and have wider function range is problematic because of high volume of calculations and vagueness in answers and CSG method has problem in terms of very limited range of volumes and drawings that it can analyze. Proposed method in this paper is an innovative method based on B-Rep in which optimization of genetic algorithm has been used to identify the relationship among the components of various views in 2D drawings. Using genetic algorithm that is a stochastic algorithm contributes that high volume of calculation that is one of main weaknesses of B-Rep method is solved. Moreover, considering correspondence condition of one to one among response in this method has caused that vagueness problem which is another weakness of B-Rep method to be almost solved so it can be said in addition to having wide range, present method doesn’t have common problems of B-Rep method and it even turns it to an effective method.

Keywords:

3D RECONSTRUCTION , ENGINEERING DRAWING , Genetic Algorithm , B-REP