Binary image interpolation based on mathematical morphology

This paper presents an interpolation technique “mmINTone” for binary images, such as logos, diagrams, graphs and cartoons. The technique removes the jaggies from a pixel-replicated image, using concepts from mathematical morphology. As a result, the edges are smoother and jaggy-free. In this paper we describe the algorithm, show some results and compare it to existing interpolation methods as well as to a previously developed technique of ours, also based on mathematical morphology. We can conclude that mmINTone is considerably faster and similar in quality.