|Advanced High Dynamic Range Imaging
Author: Francesco Banterle, Alessandro Artusi, Kurt Debattista & Alan Chalmers
High Dynamic Range photography is still a hot topic, if cooling a little because most of us know something about it or have tried it. It still fun, however, even if it doesn't always deliver on its promise of a better picture. What could be more exciting then than a book on the topic with "advanced" in the title?
Before you do get excited I need to quickly add that this is a book aimed at academics and as such the approach and presentation is very dry and the concerns go well beyond the simple task of producing a good picture.
The book starts off with a gentle introduction to HDR and the physiology, physics and math of imaging. The math starts out being used to describe fairly obvious stuff such as color spaces. It then moves on to deal with the HDR pipeline, which is just a way of outlining the stages of creation, storing, processing and displaying HDR images.
Content generation starts off with some MatLab programs to combine multiple images plus lots of math to show how the image values from a set of images can be combined to give a single HDR image. A short section on native HDR captures is mostly a look into the future or at least at equipment that is too specialized or expensive for most to use. Then the chapter moves on to consider the generation of HDR images using computer graphics and the problems of storing them. It closes with a discussion of native display of HDR images, which is interesting if a little strange and almost impractical.
Chapter 3 is dedicated to tone mapping, i.e. the reduction of the HDR range so that it can be displayed on a standard display. This in the most interesting chapter and it classifies the different approaches to tone mapping and provides some MatLab programs to implement various methods. As long as you are prepared to dig though the math and the dry academic explanations this should give you a good grasp on the sort of approaches that can be taken to tone mapping and what works in different situations.
The next chapter is of less interest because it covers the expansion of a low dynamic range image to a HDR image. This is not something most people want to do and it really only has applications in computer graphics. Chapter 5 looks at the almost separate problem of image-based lighting. The idea is that you can derive the lighting pattern from an HDR image and apply it to another image.
The final two chapters look at the problems of evaluating HDR images and compressing them for storage. The final one is interesting as it outlines and provies code for backward compatible storage using JPEG.
The book is short at around 200 pages and very dense. It isn't written in a particularly friendly style and you will have to work to extract its meaning. There are MatLab programs for most of the methods described and you can use this to clarify and experiment with the ideas.
This is not a book for the photographer unless you are working on new ways of representing HDR images - there isn't a mention of PhotoShop or a digital prosumer camera in sight. If you are planning to do any academic work on HRD or use it as part of a larger research program then it might be worth getting a copy of this book - if only to reference when you write up your work.
|Last Updated ( Wednesday, 20 July 2011 )