Survey on Video Colorization: Concepts, Methods and Applications

Leandro Stival, Hélio Pedrini

Abstract: The process of colorizing involves creating or predicting information for images or videos. While the current literature has produced relevant and high-quality results for image colorization, video colorization poses greater challenges due to additional complexities such as temporal color consistency and consistency between scenes. This work provides an analysis of the state-of-the-art techniques in video colorization, including the most commonly used colorization and machine learning techniques, as well as color space and correlated techniques to enhance current results. Unlike other reviews on colorization, this paper focuses on solutions for both natural video and animation.

External IDs:dblp:journals/vlsisp/StivalP23