Abstract
Automatic ship detection provides an essential function towards maritime domain awareness for security or economic monitoring purposes. This work presents an approach for training a deep learning ship detector in Sentinel-2 multi-spectral images with few labeled examples. We design a network architecture for detecting ships with a backbone that can be pre-trained separately. By using self supervised learning, an emerging unsupervised training procedure, we learn good features on Sentinel-2 images, without requiring labeling, to initialize our network’s backbone. The full network is then fine-tuned to learn to detect ships in challenging settings. We evaluated this approach versus pre-training on ImageNet and versus a classical image processing pipeline. We examined the impact of variations in the self-supervised learning step and we show that in the few-shot learning setting self-supervised pre-training achieves better results than ImageNet pre-training. When enough training data are available, our self-supervised approach is as good as ImageNet pre-training. We conclude that a better design of the self-supervised task and bigger non-annotated dataset sizes can lead to surpassing ImageNet pre-training performance without any annotation costs.
Highlights
Sens. 2021, 13, 4255. https://Ship detection is an important challenge in economic intelligence and maritime security, with applications in detecting piracy or illegal fishing and monitoring logistic chains.For cooperative transponders systems, such as AIS, provide ship detection and identification for maritime surveillance
For backbone pre-training with supervised learning (SSL), we look at existing large scale Sentinel 2 datasets
By varying the testing and training images we measure the transferability of the learned detector, for different levels of domain difference between training and testing sets
Summary
Ship detection is an important challenge in economic intelligence and maritime security, with applications in detecting piracy or illegal fishing and monitoring logistic chains. Cooperative transponders systems, such as AIS, provide ship detection and identification for maritime surveillance. Some ships may have non-functioning transponders; many times they are turned off on purpose to hide ship movements. Maritime patrols can help to identify suspect ships, but this requires many resources and their range is restricted. Using satellites, such as those from the European Space. Agency Sentinel-2 mission, to detect ships in littoral regions is a promising solution thanks to their large swath and high revisit time. Some commercial satellite constellations offer very high resolution images (VHR)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have