An Objective Deghosting Quality Metric for HDR Images


1Okan Tarhan Tursun, 1Ahmet Oğuz Akyüz, 2Aykut Erdem, 2Erkut Erdem

1Dept. of Comptuer Engineering, Middle East Technical University, Turkey
2Dept. of Comptuer Engineering, Hacettepe University, Turkey

Scene 1 input 1 Scene 1 input 2 Scene 1 input 3
Scene 1 Tone Mapped Result Scene 1 Metric Output
Scene 2 input 1 Scene 2 input 2 Scene 2 input 3
Scene 2 Tone Mapped Result Scene 2 Metric Output
(a) Moving people generate blending (red) and visual difference (blue) artifacts. (b) Over-smoothing gives rise to gradient inconsistency (green) artifacts.

Figure: Our metric detects several kinds of HDR deghosting artifacts. In (a), Khan et al.'s output is shown in the bottom-left corner and our metric's result in the bottom-right.
The same for (b), except Hu et al.'s deghosting algorithm is used. Exposure sequences are shown on the top. Cyan color occurs due to both gradient and visual difference metrics producing high output.

Abstract
Reconstructing high dynamic range (HDR) images of a complex scene involving moving objects and dynamic backgrounds is prone to artifacts. A large number of methods have been proposed that attempt to alleviate these artifacts, known as HDR deghosting algorithms. Currently, the quality of these algorithms are judged by subjective evaluations, which are tedious to conduct and get quickly outdated as new algorithms are proposed on a rapid basis. In this paper, we propose an objective metric which aims to simplify this process. Our metric takes a stack of input exposures and the deghosting result and produces a set of artifact maps for different types of artifacts. These artifact maps can be combined to yield a single quality score. We performed a subjective experiment involving 52 subjects and 16 different scenes to validate the agreement of our quality scores with subjective judgements and observed a concordance of almost 80%. Our metric also enables a novel application that we call as hybrid deghosting, in which the output of different deghosting algorithms are combined to obtain a superior deghosting result.