Simulation Of Enhanced Display Based On Visual Perception Of Human Eye

Computer realistic imaging based Virtual Reality (VR) technologies has extensive application value in many fields, such as military simulation training, large-scale complex equipment design, special effects in film production, digital entertainment and games, protection of cultural relics and exhibition, popular science education, urban planning, etc. However, there are many incompatible problems such as the mismatch between display technologies of existing VR systems display technologies and human eye perception. These shortcomings slow down the popularization and promotion of VR technology. In this paper, we present an enhanced display method which is based on human visual perception to improve some of these existing deficiencies. With the combination of human visual perception of psychology, physiology knowledge and high dynamic range image processing technology, we simulate the light and dark adaptation dynamic process and gaze effect of human eye.In chapter one, we introduce the development of virtual reality technology, its application areas and present status. Then we point out its current problems as well as outline our main research work and the full text organization.The second chapter describes the related basic knowledge of physiological and psychological theory concerning the human visual system, which provides a basis for our algorithm design and establishment.In Chapter three, we propose a new light and dark adaptation simulation method which is based on intensity-response model grounded on human eye response mechanism. According to different physiological response mechanism of cone and rod cells, this algorithm computes vary intensity response under changing scene luminance. By combining with high dynamic range image processing technology and mesopic factor from experimental psychology, we can simulate the adaptation visual process in mesopic vision which is between photopic vision and scotopic vision. Inverse response transform and chromatic adaptation transform is applied to obtain the final luminance, in order to simulate the human eye visual perception in the dynamic process of light and dark adaptation in virtual 3D scene.In chapter four, we suggest an efficient Foveation Filtering algorithm based on Mipmap Pyramid to simulate the gaze effect of human eye. At first the spatial-variant Gaussian kernel is calculated according to critical frequency; then we relate it to corresponding Mipmap level by using pixel-wise method of Root-Mean-Square-Error (RMSE) to find a best match; due to the nonlinear Mipmap interpolation under Bilateral Filtering scheme, we obtain the final color for display to simulate the gaze effect of human eye in realtime.This work is summarized at the last chapter, as well as the prospect of future research.