Practical Auto-Calibration for Spatial Scene-Understanding from Crowdsourced Dashcamera Videos

TL;DR

This paper presents a practical system for auto-calibrating monocular dashcam videos from crowdsourced data, enabling accurate depth and ego-motion estimation without prior camera calibration.

Contribution

It introduces a novel auto-calibration method that handles uncalibrated crowdsourced videos, improving scene understanding for autonomous driving applications.

Findings

Effective calibration on KITTI, Oxford RobotCar, and D$^2$-City datasets.

Enables accurate monocular dense depth and ego-motion estimation.

Handles challenging motion sequences with reconstruction ambiguities.

Abstract

Spatial scene-understanding, including dense depth and ego-motion estimation, is an important problem in computer vision for autonomous vehicles and advanced driver assistance systems. Thus, it is beneficial to design perception modules that can utilize crowdsourced videos collected from arbitrary vehicular onboard or dashboard cameras. However, the intrinsic parameters corresponding to such cameras are often unknown or change over time. Typical manual calibration approaches require objects such as a chessboard or additional scene-specific information. On the other hand, automatic camera calibration does not have such requirements. Yet, the automatic calibration of dashboard cameras is challenging as forward and planar navigation results in critical motion sequences with reconstruction ambiguities. Structure reconstruction of complete visual-sequences that may contain tens of thousands of images is also computationally untenable. Here, we propose a system for practical monocular onboard camera auto-calibration from crowdsourced videos. We show the effectiveness of our proposed system on the KITTI raw, Oxford RobotCar, and the crowdsourced D$^2$-City datasets in varying conditions. Finally, we demonstrate its application for accurate monocular dense depth and ego-motion estimation on uncalibrated videos.