A Review on Example-based Colourization

Abstract

Pseudo colouring is the process of adding colour to a gray scale image in order to highlight specific features in an image that are unnoticeable in the gray scale version. This process can also be applied to recolour an existing coloured image. In past few years, several image colourization techniques have been developed to improve quality of colourization. From which most efficient and robust technique is example-based colourization. Example-based method colourizes a gray scale image with the help of a reference colour image. This method produce visually plausible result if there exist a strong correlation between reference image and target image. Usually reference image is provided by user or it can be automatically retrieved using image retrieval system. This paper provide a review on several example-based methods. Different and recent example-based methods like example-based colourization using normalized cross correlation, locality consistent sparse representation, automatic feature selection and fusion, location aware cross scale matching, variational colour transfer and descriptor based colourization are reviewed in this paper.