Extra Colloquium Computer Science, Prof. Kenneth Camilleri (University of Malta)

Title: Circle-Based Vectorisation (CBV): Vectorising Drawings having Shadow Cues

                           

Abstract:

The creation of virtual 3D models from two-dimensional sketches allows a professional
designer to rapidly obtain 3D visualisations or physical prototypes of their concept sketches,
and opens up object design to lay people not having CAD experience.
A simple drawing may have a large number of possible 3D interpretations, of which only a
small subset is physically realisable and thus deemed plausible by human observers. Edge
boundary strokes are sufficient to elicit a 3D interpretation of a sketched object but designers
often introduce artistic cues to assist the observer in arriving to the intended form
interpretation. Therefore, exploiting such artistic cues allows for a more accurate and
efficient machine interpretation of form drawings. Shadows are a common type of artistic cue
that is used to modulate the interpretation of the drawing. However, vectorisation methods –
converting a raster image into vectors, resulting in the constituent lines of the drawing – are
not designed to handle such cues.

This talk will describe our novel circle-based vectorisation (CBV) algorithm that is robust to
shadow cues, inherently parallel, sparse and scalable, implicitly extracting the drawing
junctions. CBV accumulates evidence from families of concentric sampling circles to detect
junctions and lines without requiring any specific centering of the circle samplers, extending
he state-of-the-art drawing vectorisation which require cue-less line drawings and centred
samplers. Furthermore, CBV may exploit the obtained graph to detect missing-line errors,
and the 3D geometry to detect incorrect junctions, triggering a reduced-scale vectorisation at
the error locations.

Colloquium coordinators are Prof.dr. M. Aiello (e-mail : M.Aiello rug.nl ) and
Prof.dr. M. Biehl (e-mail: M.Biehl rug.nl )

http://www.rug.nl/research/jbi/news/colloquia/computerscience