Camera response function estimation and application with a single image

Abstract

Camera response function (CRF) relates scene irradiance to image intensities. Estimation of CRF is a fundamental and necessary step in many computer vision applications such as the generation of high dynamic image, bidirectional reflectance distribution function (BRDF) estimation etc. In the paper, we compute CRF from a color image and edit images based on CRF to enhance contrast and change exposure of images. This method selects suitable edge windows with two uniform color regions, and is based on the empirical prior on camera response functions. The method uses only one color image rather than a set of images. The experiments show our estimation result is accurate. © 2011 Springer-Verlag.