Automated hedcut illustration using isophotes

Abstract

In this work, we present an automated system for creating hedcut illustrations, portraits rendered using small image feature aligned dots (stipples). We utilize edge detection and shading cues from the input photograph to direct stipple placement within the image. Both image edges and isophotes are extracted as a means of describing the image feature and shading information. Edge features and isophotes are then assigned different priorities, with isophotes being assigned the highest priority to enhance the depth perception within the hedcut portrait. Priority assignment dictates the stipple alignment and spacing. Finally, stipple size is based on the number of points and intensity and the gradient magnitude of the input image. © 2010 Springer-Verlag Berlin Heidelberg.