Image Appearance Modeling and High-Dynamic-Range Image Rendering

摘要

Recently, the field of color appearance modeling has been extended further into the spatial and temporal domains through efforts known as image appearance modeling. Image appearance models have applications in rendering and visualization of image data, cross-media image and video reproduction, and image quality specification. This presentation provides an overview of one such model, iCAM, illustrates several examples of its application, and reviews a psychophysical experiment aimed at evaluating iCAM and other algorithms for tone mapping of high-dynamic-range (HDR) images. Initial psychophysical results for HDR image rendering indicates that the Durand and Dorsey [2002] algorithm produces the most preferred results in a general sense. A full technical report on this work can be found at < www.cis.rit.edu/mcsl/research/reports.html >. Example iCAM code and further references can be found at < www.cis.rit.edu/mcsl/iCAM >.