Marine vessel tracking using a monocular camera

TL;DR

This paper introduces a real-time, low-power monocular camera-based method for tracking marine vessels, utilizing GPS data for calibration and bounding box metrics for distance estimation, suitable for IoT environments.

Contribution

It presents a novel approach combining GPS-based camera calibration with bounding box analysis for vessel tracking from monocular video.

Findings

Average prediction error of 5.55m per 100m distance

Real-time processing capability on low-powered IoT devices

Effective multi-vessel tracking in maritime environments

Abstract

In this paper, a new technique for camera calibration using only GPS data is presented. A new way of tracking objects that move on a plane in a video is achieved by using the location and size of the bounding box to estimate the distance, achieving an average prediction error of 5.55m per 100m distance from the camera. This solution can be run in real-time at the edge, achieving efficient inference in a low-powered IoT environment while also being able to track multiple different vessels.