Perspective Fields for Single Image Camera Calibration

TL;DR

This paper introduces Perspective Fields, a novel image representation capturing local perspective properties that are robust to editing and can be used for camera calibration and image compositing.

Contribution

It proposes Perspective Fields as a minimal-assumption, interpretable representation for local perspective, trained via neural networks for camera calibration tasks.

Findings

Robustness under various editing scenarios

Effective camera calibration from Perspective Fields

Improved image compositing applications

Abstract

Geometric camera calibration is often required for applications that understand the perspective of the image. We propose perspective fields as a representation that models the local perspective properties of an image. Perspective Fields contain per-pixel information about the camera view, parameterized as an up vector and a latitude value. This representation has a number of advantages as it makes minimal assumptions about the camera model and is invariant or equivariant to common image editing operations like cropping, warping, and rotation. It is also more interpretable and aligned with human perception. We train a neural network to predict Perspective Fields and the predicted Perspective Fields can be converted to calibration parameters easily. We demonstrate the robustness of our approach under various scenarios compared with camera calibration-based methods and show example applications in image compositing.