The Role of Image Statistics on Cross-Media Gamut Mapping

摘要

A universal gamut mapping algorithm (GMA) which performs well for any images in color reproduction between different media would be highly desirable. However, various GMAs have previously been reported to have image-dependent behavior. In order to deal with this problem, recent studies focus on image adapted solutions. Nevertheless, these solutions normally were developed based on a predetermined image characteristic which is incapable of optimizing all types of reproductions. The present study hence investigates various image statistics and their related GMA performance for a better understanding of their correlation. In this paper, the process to identify important image statistics is proposed and an automatic approach to choose the best-suited GMAs for a certain image is also demonstrated.