D-VPnet: A Network for Real-time Dominant Vanishing Point Detection in Natural Scenes

TL;DR

This paper introduces D-VPnet, a CNN-based method for real-time detection of dominant vanishing points in natural scenes, improving accuracy and speed over existing techniques by utilizing feature line proposals and parallel line cues.

Contribution

The paper presents a novel CNN architecture with a feature line-segment proposal unit for efficient, accurate, real-time dominant vanishing point detection in outdoor environments.

Findings

Outperforms state-of-the-art methods in accuracy.

Achieves real-time detection at 115fps.

Validated on public and specialized datasets.

Abstract

As an important part of linear perspective, vanishing points (VPs) provide useful clues for mapping objects from 2D photos to 3D space. Existing methods are mainly focused on extracting structural features such as lines or contours and then clustering these features to detect VPs. However, these techniques suffer from ambiguous information due to the large number of line segments and contours detected in outdoor environments. In this paper, we present a new convolutional neural network (CNN) to detect dominant VPs in natural scenes, i.e., the Dominant Vanishing Point detection Network (D-VPnet). The key component of our method is the feature line-segment proposal unit (FLPU), which can be directly utilized to predict the location of the dominant VP. Moreover, the model also uses the two main parallel lines as an assistant to determine the position of the dominant VP. The proposed method was tested using a public dataset and a Parallel Line based Vanishing Point (PLVP) dataset. The experimental results suggest that the detection accuracy of our approach outperforms those of state-of-the-art methods under various conditions in real-time, achieving rates of 115fps.