Hand-drawn Grayscale Image Colorful Colorization Based on Natural Image
Liyang Fang,Jin Wang,Guodong Lu,Dongliang Zhang,Jianhui Fu
DOI: https://doi.org/10.1007/s00371-018-1613-8
2018-01-01
Abstract:This paper presents a novel colorization technique for hand-drawn grayscale images, such as cartoons and sketches, based on reference natural images and simple user interactions, which can generate colorful and natural results. Firstly, to solve the problem of inputting complex color scribbles by simple interactions, we introduce a region thinning method for generating color scribbles and then map these scribbles to the corresponding regions in the hand-drawn image via region mapping method based on the scanning line. Secondly, to maintain the color fidelity, a luminance downward modification method is proposed. Next, a optimization method for colorization is proposed, where the feature vectors are adjusted by a smooth feature map, thus maintaining smooth color transitions in the smooth areas and controlling the color overflow in the texture areas. Thirdly, to fuse the colorized image with a highlight effect, a luminance upward modification method by weights, which are determined by color distance and boundary distance, is proposed. The experimental results of the proposed algorithm show the smooth color transition in the intensity-continuity areas, color overflow controlling in the texture-continuity areas and natural effect of light and shadow.