Detecting mirror symmetry in single-view wireframe sketches of polyhedral shapes. (October 2016)

• Record Type: Journal Article
• Authors: Plumed, Raquel; Company, Pedro; Varley, Peter A.C.
• Is Part Of: Computers & graphics. Volume 59(2016)
• Page Start: 1
• Page End: 12
• Publication Date: 2016-10
• Subjects: Sketch-based modelling -- Artificial perception -- Bilateral symmetry; Computer graphics -- Periodicals; 006.6
• DOI: 10.1016/j.cag.2016.05.013
• Languages: English
• ISSNs: 0097-8493

Abstract

Abstract: As part of a strategy for creating 3D models of engineering objects from a sketched input, we attempt to identify mirror symmetry planes early in the process. Our input is a 2D line-drawing derived from a single view sketch of a polyhedral shape, and suitable cues are used to identify mirror symmetry planes algorithmically. Previous attempts to simplify the search space use incomplete criteria. In this paper, we further simplify the search space by scoring valid symmetry planes by their likelihood, i.e., we do not assert "this is a symmetry plane" but "this is more/less likely to be a symmetry plane". Abstract : Graphical abstract: Our input is a 2D line-drawing derived from a single view sketch of a polyhedral shape, and suitable cues are used to identify mirror symmetry planes algorithmically. We do not assert "this is a symmetry plane" but "this is more/less likely to be a symmetry plane." Abstract : Highlights: We identify mirror symmetry planes in single view sketches of polyhedral shapes. The approach uses suitable cues to identify algorithmically mirror symmetry planes. The approach improves as fast as it considerably reduces the search space. The approach scores symmetry planes by their likelihood.