Frontal geometry from sketches of engineering objects: is line labelling necessary?

Abstract

A tool which can quickly interpret line drawings (with hidden lines removed) of engineering objects as boundary representation CAD models would be of significant benefit in the process of engineering design. Inflation of the drawing to produce a frontal geometry, a geometric realisation of that part of the object visible in the drawing, is an important stage of this process. Previous methods of producing frontal geometries have relied on the technique of line labelling (labelling edges as convex, concave or occluding). Although restricted subsets of the line-labelling problem have known solutions, reliable methods have not been found for the general line-labelling problem, and traditional methods, when adapted to drawings with non-trihedral junctions, are unacceptably slow. Many other papers assume that line labelling is an essential step. Here, we show this is not necessarily true, and that comparable results can be obtained by a novel alternative approach. Firstly, we consider what outputs from line labelling are essential to the production of frontal geometry. Secondly, we investigate by what other means these outputs can be produced. Our work indicates that the only essential output from line labelling for frontal geometry is the determination of which T-junctions in a drawing are occluding and which are non-occluding. This information is required for inflation, and also for detection of symmetry and for constructing hidden topology. Thus, we propose and analyse a new method which, in the absence of line labels, simultaneously inflates a drawing to produce the frontal geometry and attempts to determine whether each T-junction is occluding or not. For drawings of objects with holes or pockets, and for cases where line labelling is particularly unreliable, our new method can provide a better alternative.