On Localizing a Camera from a Single Image

TL;DR

This paper demonstrates that combining projective geometry, neural networks, and crowd-sourced data enables precise localization of cameras from single images, significantly outperforming previous neural network methods.

Contribution

It introduces a novel approach that integrates geometry and machine learning to accurately estimate camera locations from single images, achieving high precision.

Findings

95% of images localized within 12 meters

Outperforms PoseNet by two orders of magnitude

Enables design of virtual sensors with accurate parameters

Abstract

Public cameras often have limited metadata describing their attributes. A key missing attribute is the precise location of the camera, using which it is possible to precisely pinpoint the location of events seen in the camera. In this paper, we explore the following question: under what conditions is it possible to estimate the location of a camera from a single image taken by the camera? We show that, using a judicious combination of projective geometry, neural networks, and crowd-sourced annotations from human workers, it is possible to position 95% of the images in our test data set to within 12 m. This performance is two orders of magnitude better than PoseNet, a state-of-the-art neural network that, when trained on a large corpus of images in an area, can estimate the pose of a single image. Finally, we show that the camera's inferred position and intrinsic parameters can help design a number of virtual sensors, all of which are reasonably accurate.