Interactive shadow analysis for camera heading in outdoor images

Image geo-localization is an important problem with many applications such as augmented reality and navigation. The most common ways to geo-localize an image are to use its meta-data such as GPS or to match it against a geotagged database. When neither of those is available, it is still possible to apply shadow analysis to determine the camera heading for outdoor images. This could be useful pruning the search space in geo-localization applications, for example by removing roads with incompatible orientations from a database such as Open Street Map. In this paper, we develop a novel interactive method for deducing the global heading of a query image using the shadows in it. We start by constructing a model of the sun-earth system to determine all shadows possible at a given approximate latitude, and compare shadows within the query to those possible under the model to determine the range of possible headings. We demonstrate this on 54 query images with known ground truth, and show that in 52 cases the ground truth lies in the computed range.

1. Introduction

Image geo-localization is an important problem with many applications such as augmented reality and navigation. The easiest way to geo-localize an image is to use the meta-data from sensors integrated with the imaging device such as GPS or compass in order to determine the location and heading of the captured imagery. However, many images lack such meta-data, or have erroneous meta-data due to various physical conditions. For example, GPS is known to be erroneous in urban canyon areas, can at times be jammed by adversarial forces, and is unavailable indoors. Compass data, which relies on detection of Earth's magnetic field, is unreliable in the presence of power transmission lines which generate their own magnetic fields and steel cabinets which disturb existing magnetic fields [1].