Monocular 3D Vehicle Detection Using Uncalibrated Traffic Cameras through Homography

TL;DR

This paper introduces a novel method for 3D vehicle detection using uncalibrated monocular traffic cameras by leveraging homography estimation and a dual-view network to generate accurate bird's eye view detections.

Contribution

It presents a new approach that estimates homography without camera calibration and uses a dual-view network with tailed r-box regression for improved detection accuracy.

Findings

Effective 3D vehicle detection from uncalibrated cameras

Generalizes well to new environments and camera setups

Outperforms existing methods in accuracy

Abstract

This paper proposes a method to extract the position and pose of vehicles in the 3D world from a single traffic camera. Most previous monocular 3D vehicle detection algorithms focused on cameras on vehicles from the perspective of a driver, and assumed known intrinsic and extrinsic calibration. On the contrary, this paper focuses on the same task using uncalibrated monocular traffic cameras. We observe that the homography between the road plane and the image plane is essential to 3D vehicle detection and the data synthesis for this task, and the homography can be estimated without the camera intrinsics and extrinsics. We conduct 3D vehicle detection by estimating the rotated bounding boxes (r-boxes) in the bird's eye view (BEV) images generated from inverse perspective mapping. We propose a new regression target called tailed r-box and a dual-view network architecture which boosts the detection accuracy on warped BEV images. Experiments show that the proposed method can generalize to new camera and environment setups despite not seeing imaged from them during training.