TR-IIS-08-006    Fulltext

Abstract

In this paper, we propose an effective scheme for enhancing the visual details of digital images with the minimal amount of user adjustment. Digital archives are becoming increasingly popular due to the development of convenient and powerful digitizing techniques. However, a substantial number of digital images suffer from loss of detail because they were captured with old-fashioned equipment. Thus, an automatic tone reproduction technique is needed. We attempt to solve the above issues by combining a novel local normalization concept with an adaptive contrast assessment process. The proposed tone reproduction scheme effectively enhances poor quality regions, while simultaneously preserving good quality regions with default parameter settings. As the proposed scheme eliminates most of the manual effort required to adjust the parameters, it can be considered as nearly automatic. The results of experiments demonstrate that the scheme outperforms many existing algorithms when applied to restoring digital images for a national digital archive program.

index terms: Detail preserving, tone reproduction