Automatic Detection of Dark Ship-to-Ship Transfers using Deep Learning and Satellite Imagery

TL;DR

This work addresses the challenge of detecting dark Ship-to-Ship transfers that evade AIS monitoring by fusing AIS-derived spatiotemporal localization with optical satellite imagery. It combines Segment Anything Model segmentation and YOLOv8m object detection to identify STS in PlanetScope imagery, then cross-references with AIS to flag transfers lacking sufficient AIS signatures. The approach identifies 402 dark STS in the Kerch Strait between 2021 and September 2023, demonstrating a lightweight detector capable of near real-time analysis and highlighting the method's potential for improving maritime security and sanctions enforcement. The study advances transshipment surveillance by integrating multi-sensor data and deep learning to reveal illicit activity that AIS alone cannot capture.

Abstract

Despite extensive research into ship detection via remote sensing, no studies identify ship-to-ship transfers in satellite imagery. Given the importance of transshipment in illicit shipping practices, this is a significant gap. In what follows, I train a convolutional neural network to accurately detect 4 different types of cargo vessel and two different types of Ship-to-Ship transfer in PlanetScope satellite imagery. I then elaborate a pipeline for the automatic detection of suspected illicit ship-to-ship transfers by cross-referencing satellite detections with vessel borne GPS data. Finally, I apply this method to the Kerch Strait between Ukraine and Russia to identify over 400 dark transshipment events since 2022.

Figures (6)

• Figure 1: A sample of automatically detected dark STS
• Figure 2: Dark Ship-to-Ship Transfer Detection Pipeline
• Figure 3: Cross-referencing AIS and Satellite Imagery
• Figure 4: YOLOv8m model diagnostics
• Figure 5: Normalized class-wise confusion matrix