In this paper, we present work in progress towards the generation of a ground truth data set for High Dynamic Range (HDR) stereo matching. The development and evaluation of novel stereo matching algorithms that are tailored to the characteristics of HDR images would greatly benefit from the availability of such reference data. We describe our laboratory setup and processing steps for acquiring multi-exposed stereo images along with a corresponding reference disparity map computed by a structured light approach. We discuss the special requirements and challenges which HDR test scenes impose on the computation of the reference disparities and show some preliminary results.
T. Akhavan, C. Kapeller, J. Cho, M. Gelautz: "Stereo HDR Disparity Map Computation Using Structured Light"; Talk: HDRi2014, Sarajevo; 03-04-2014; in: "HDRi2014 - Second International Conference and SME Workshop on HDR imaging", (2014), 4 pages.
Click into the text area and press Ctrl+A/Ctrl+C or ⌘+A/⌘+C to copy the BibTeX into your clipboard… or download the BibTeX.