A simplified HDR image processing pipeline for digital photography

Singnoo, Jakkarin (2012) A simplified HDR image processing pipeline for digital photography. Doctoral thesis, University of East Anglia .

Preview

PDF Download (9MB) | Preview

Abstract

High Dynamic Range (HDR) imaging has revolutionized the digital imaging. It allows
capture, storage, manipulation, and display of full dynamic range of the captured scene.
As a result, it has spawned whole new possibilities for digital photography, from photorealistic
to hyper-real. With all these advantages, the technique is expected to replace
the conventional 8-bit Low Dynamic Range (LDR) imaging in the future. However,
HDR results in an even more complex imaging pipeline including new techniques for
capturing, encoding, and displaying images. The goal of this thesis is to bridge the
gap between conventional imaging pipeline to the HDR’s in as simple a way as possible.
We make three contributions. First we show that a simple extension of gamma
encoding suffices as a representation to store HDR images. Second, gamma as a control
for image contrast can be ‘optimally’ tuned on a per image basis. Lastly, we show
a general tone curve, with detail preservation, suffices to tone map an image (there is
only a limited need for the expensive spatially varying tone mappers). All three of our
contributions are evaluated psychophysically. Together they support our general thesis
that an HDR workflow, similar to that already used in photography, might be used. This
said, we believe the adoption of HDR into photography is, perhaps, less difficult than it
is sometimes posed to be.

Item Type:	Thesis (Doctoral)
Faculty \ School:	Faculty of Science > School of Computing Sciences
Depositing User:	Zoe White
Date Deposited:	07 Feb 2014 10:31
Last Modified:	07 Feb 2014 10:31
URI:	https://ueaeprints.uea.ac.uk/id/eprint/41945
DOI:

Downloads

Actions (login required)

View Item