Detecting Vanishing Points using Global Image Context in a Non-Manhattan World

TL;DR

This paper introduces a deep learning-based method for detecting horizontal and zenith vanishing points in non-Manhattan environments, leveraging global image context to improve accuracy and speed.

Contribution

It reverses traditional vanishing point detection by scoring horizon line candidates with global context, avoiding Manhattan assumptions, and achieving state-of-the-art results.

Findings

State-of-the-art accuracy on three benchmarks

Significantly faster than previous methods

Effective in scenes with a single horizontal vanishing point

Abstract

We propose a novel method for detecting horizontal vanishing points and the zenith vanishing point in man-made environments. The dominant trend in existing methods is to first find candidate vanishing points, then remove outliers by enforcing mutual orthogonality. Our method reverses this process: we propose a set of horizon line candidates and score each based on the vanishing points it contains. A key element of our approach is the use of global image context, extracted with a deep convolutional network, to constrain the set of candidates under consideration. Our method does not make a Manhattan-world assumption and can operate effectively on scenes with only a single horizontal vanishing point. We evaluate our approach on three benchmark datasets and achieve state-of-the-art performance on each. In addition, our approach is significantly faster than the previous best method.