Lighting Estimation in Outdoor Image Collections
3D International Conference on 3D Vision (3DV) 2014
December 8, 2014
Large scale structure-from-motion (SfM) algorithms have recently enabled us to reconstruct highly detailed 3-D models of our surroundings simply by taking photographs. In this paper, we propose to leverage these reconstruction techniques and automatically estimate the outdoor illumination conditions for each image in a SfM photo collection. First, we introduce a novel dataset of outdoor photo collections, where the ground truth lighting conditions are known at each image. Second, we present an inverse rendering approach that recovers a high dynamic range estimate of the lighting conditions for each low dynamic range input image. Finally, we use our novel database to quantitatively evaluate the performance of our algorithm. Results show that lighting can be faithfully estimated, both in terms of light direction and intensity.
Download File "Lighting Estimation in Outdoor Image Collections-Paper"
[pdf, 18.55 MB]