Interactive Image Colorization and Recoloring based on Coupled Map Lattices

Abstract

Colorization is a computer-assisted process of adding color to a monochrome image or movie. Most current colorization algorithms either require a significant user effort or have large computational time. In any case, colorization of real size images remains a time-consuming, tedious task. In this paper we present a new colorization method, based on GrowCut image segmentation algorithm. In our approach a user just marks some pixels with desired colors, and the algorithm propagates these colors to the remainder of the image. After the initial colorization is computed, the user can interactively adjust and refine the colors of the image. The main advantage of the proposed method over the existing ones is that modification and refining of the image colors doesn’t lead to full recomputation of the image colorization and is performed very fast, thanks to evolutionary nature of the coupled map lattices (CML). Our algorithm can also be successfully applied to image recoloring. We demonstrate our results on a variety of examples.