Online sketch recognition using geometric-based classifiers

Shuxia Wang, Guanfeng Wang, Sheng-feng Qin

Research output: Contribution to conferencePaperpeer-review

2 Citations (Scopus)

Abstract

This paper proposes a novel method using geometric-based classifiers, to recognize stroke as geometric primitives and develops a human-computer interface prototype system (FSR) for assisting designers during conceptual design stages, which make system interface easy and friendly to use. The system is designed to infer designers' intention and recognizes the input single strokes into more 2D geometric primitives including line-segments, polylines, circles, circular arcs, ellipses, elliptical arcs, hyperbolas and parabolas. The geometric features are invariant with rotation of figures, including polygonal approximation points, enclosing rectangle, least medium squares error. The filters and fuzzy classifiers are built from these geometric features. The human interaction can help to determine or revise the ambiguous results or misrecognitions. The test results showed that the proposed method can support freehand sketching based conceptual design with a satisfactory interpretation rate.
Original languageEnglish
Publication statusPublished - Sep 2012
Event2012 18th International Conference on Automation and Computing (ICAC) - Loughborough
Duration: 1 Sep 2012 → …

Conference

Conference2012 18th International Conference on Automation and Computing (ICAC)
Period1/09/12 → …

Fingerprint Dive into the research topics of 'Online sketch recognition using geometric-based classifiers'. Together they form a unique fingerprint.

Cite this