Understanding Folding Process of Origami Drill Books Based on Graphic Matching

This paper introduces an approach for recognizing illustrations of origami drill books with a view to understanding folding process of origami. For this purpose, comparing and matching two kinds of graphic forms are required. One is a graph generated from an origami illustration image by applying image processing techniques to it. The other is that obtained from an internal model which is produced and updated dynamically according to the results of understanding of the folding process stepwise. In order to comparing and matching these graphics, we propose a method for extracting USGs from origami illustration images. This method is based on Hough transform and takes into account the characteristics of origami illustration images. We also propose a method for matching two kinds of graphic forms based on a genetic algorithm. Experiments show that our methods are effective in recognizing illustration images including a lot of noise.