A New Photoreceptor Model of Human Visual System for HDR Natural Luminance Perception

Abstract

The human visual system is a very complex system and the photoreceptors on the retina are mainly responsible for sensing the luminance of natural scenes. In this paper, a new model which can simulate the adaptive luminance characteristics of the human eye’s photoreceptor cells is proposed. According to this model, a new tone mapping algorithm which can display high dynamic range (HDR) images on low dynamic range (LDR) display devices is presented. This technology make the final mapped image simulate the viewer’s photorealistic in real scene. First, a HDR image was divided into bright, middle and dark regions, and the three corresponding response curves for each region were acquired. Then, each curve applied different weighting factors, which was obtained by counting the number of pixels within each region. Finally, the final tone mapping curve was synthesized. The proposed algorithm which does not require manually setting parameters has strong stability and it is easy to use. In a variety of experimental conditions for natural scenes, the proposed algorithm has a strong realism. The proposed algorithm was also compared with existing major tone mapping algorithm based on traditional photoreceptor model.