High Dynamic Range Image Tone Mapping with Edge Preserved Dithering

Abstract

Current tone mapping technique for high dynamic range (HDR) image is limited by the gray scale of display screen. In this paper, we propose a novel tone mapping approach based on edge preserved dithering, which makes use of high resolution of display screen to breakthrough the gray scale limitation. To each HDR image, we first decompose its L* component to the ditherable patches and the unditherable patches. Then, we utilize linear programming to tone map the mean gray levels of the ditherable patches and the gray level of each pixel in the unditherable patches to an extended dynamic range. Finally, the gray levels of the tone mapped image are further narrowed to the required gray scale of display screen in which dithering is applied on ditherable patches. Experimental results show that edge preserved dithering can improve the distribution of the limited gray levels in the generated images and effectively reduce contouring and detail loss.