LinDA：Feature Extraction and Classification of Line Segments for Semi-Automatic Generation of Cartoon Background Images

Abstract

For cartoon novices, it is effective to reuse as their background images, line drawing materials drawn and opened to the public by other artists. Due to the versatility, however, such materials do not fit into a specific user’s cartoon with his/her own handwritten features. In this study, therefore, we propose a method to semi-automatically convert vector drawings composed of line segments with uniform thickness into those with an individual style. We have constructed a system that implements this method—LinDA（Line Drawing Artist）, with a primary objective to generate totally consistent cartoons in an efficient manner.