Reasoning about Photo Collections using Models of Outdoor Illumination

Daniel Hauagge, Scott Wehrwein, Paul Upchurch, Kavita Bala, Noah Snavely

Natural illumination from the sun and sky plays a significant role in the appearance of outdoor scenes. We propose the use of sophisticated outdoor illumination models, developed in the computer graphics community, for estimating appearance and timestamps from a large set of uncalibrated images of an outdoor scene. We first present an analysis of the relationship between these illumination models and the geolocation, time, surface orientation, and local visibility at a scene point. We then use this relationship to devise a data-driven method for estimating per-point albedo and local visibility information from a set of Internet photos taken under varying, unknown illuminations. Our approach significantly extends prior work on appearance estimation to work with sun-sky models, and enables new applications, such as computing timestamps for individual photos using shading information.


Full Paper BMVC2014
PDF (23MB)
Poster presented at BMVC2014
PDF (35MB)

BibTeX entry

   Title = {Reasoning about Photo Collections using Models of Outdoor Illumination},
   Author = {Daniel Hauagge and Scott Wehrwein and Paul Upchurch
             and Kavita Bala and Noah Snavely},
   booktitle = {Proceedings of BMVC},
   Year = {2014}

Acknowledgments.This work was supported in part by NSF grants IIS-0964027, IIS-1111534, IIS-1149393, by the Intel Science and Technology Center for Visual Computing, and by the Sloan Foundation. We are grateful to Daniel Carpenter for the ground truth Castle timestamps.