High Dynamic Range Imaging

Erik Reinhard
School of Computer Science
University of Central Florida, Orlando


Conventional images are represented with a byte per color channel per pixel. This is not sufficient to represent scenes accurately. High dynamic range (HDR) images ameliorate this problem by essentially representing each pixel as a triplet of floating point numbers. However, such images cannot be displayed directly on conventional display devices. To prepare HDR images for display, an extra step is required - called tone reproduction or tone mapping. While several tone reproduction algorithms exist, choosing an appropriate one for a specific task remains difficult.

After a general introduction to tone reproduction, in this presentation two specialized topics relating to tone reproduction are discussed, namely a proposal for validation of tone reproduction algorithms, and the use of color appearance models in high dynamic range imaging.


Erik Reinhard is assistant professor at the University of Central Florida and has an interest in the fields of visual perception and parallel graphics. He has a B.Sc. and a TWAIO diploma in computer science from Delft University of Technology, and a Ph.D. in computer science from the University of Bristol. He was a post-doctoral researcher at the University of Utah. He is founder and editor-in-chief of ACM Transactions on Applied Perception, and is also writing a book on high dynamic range imaging which will be published in the summer of 2005 by Morgan Kaufmann.

Colloquia Series page.