Online Self-Supervised Thermal Water Segmentation for Aerial Vehicles
This work addresses the challenge of autonomous aerial robots operating in near-shore environments at night by providing a thermal water segmentation capability, though it is incremental as it adapts existing methods to a new domain.
The paper tackles the problem of scarce thermal data for water segmentation in aerial vehicles by introducing an online self-supervised method that adapts an RGB-trained network to thermal imagery using texture and motion cues, enabling tasks like navigation and flow tracking at night and outperforming supervised models on limited data.
We present a new method to adapt an RGB-trained water segmentation network to target-domain aerial thermal imagery using online self-supervision by leveraging texture and motion cues as supervisory signals. This new thermal capability enables current autonomous aerial robots operating in near-shore environments to perform tasks such as visual navigation, bathymetry, and flow tracking at night. Our method overcomes the problem of scarce and difficult-to-obtain near-shore thermal data that prevents the application of conventional supervised and unsupervised methods. In this work, we curate the first aerial thermal near-shore dataset, show that our approach outperforms fully-supervised segmentation models trained on limited target-domain thermal data, and demonstrate real-time capabilities onboard an Nvidia Jetson embedded computing platform. Code and datasets used in this work will be available at: https://github.com/connorlee77/uav-thermal-water-segmentation.