Abstract. Effective maritime surveillance and small-scale fisheries management remain challenging in coastal waters, where small vessels are not systematically tracked and are often poorly represented in medium-resolution satellite imagery. Within the AI4COPSEC Horizon Europe framework, this study investigates an object-detection workflow for monitoring small vessels along the Adriatic coasts of Marche and Puglia, Italy. Sentinel-2 and high-resolution PlanetScope RGB imagery were manually annotated to build a task-specific optical dataset and to fine-tune models previously pretrained on a larger SAR-optical vessel dataset. This two-stage strategy was designed to exploit heterogeneous vessel representations during pretraining and then adapt the detector to the target coastal optical domain. The resulting dataset comprised 4,202 image tiles for pretraining and 706 tiles for fine-tuning, with 16,096 and 1,716 vessel annotations, respectively, all belonging to a single target class. Detection experiments were conducted using several YOLOv26 configurations trained under a consistent protocol to assess the trade-off between model complexity, accuracy, and computational efficiency. Among the standard variants, YOLOv26-M achieved the most balanced performance, with Precision of 0.813, Recall of 0.846, F1-score of 0.829, Accuracy of 0.719 and mAP50-95 of 0.306. Pruned and lightweight alternatives showed competitive efficiency-oriented behaviour. Results indicate that, in small-target coastal environments, increasing model size does not necessarily yield proportional gains, whereas task-oriented architectural design improves the balance between detection quality and computational cost.
S. Chiappini, A. Galdelli, Andrea Fiorani et al.· The International Archives o...· 0 citations
Abstract. Extracting road networks from high-resolution imagery remains a significant challenge in geomatics, particularly in fragmented rural landscapes. The big difficulty is the spectral similarities between unpaved tracks and agricultural backgrounds that can lead to classification errors. This study proposes an automated geospatial pipeline based on the YOLOv11 architecture. Specifically, the approach is made on the optimization of the multi-class road detection in the rural areas of Kosina and Markowa, two villages in Poland. To reduce the computational effort, due to large-scale 9000×9000 px orthophotos and to improve the detection of small-scale features, Slicing Aided Hyper Inference (SAHI) strategy was integrated. High-resolution imagery has been decomposed into optimized tiles, ensuring feature continuity across boundaries and preventing GPU memory overhead. The instance segmentation model was trained on a custom-annotated dataset, with seven labels (categories) such as internal paved roads, rural tracks, and railway infrastructures. Therefore, a high level of robustness has been achieved reaching a mean Average Precision value (mAP@0.5) of 0.90. A confusion matrix reveals quantitatively that the pipeline effectively distinguishes between complex classes and low omission rates. As a result, the generated outputs are converted into interoperable GeoJSON format ensuring their integration into GIS environments. In conclusion, the experimental result demonstrates that the framework is valuable for emergency response logistics and urban planning. It offers a scalable and near real-time solution for updating national topographic databases.
M. Sanità, L. Nepi, E. Malinverni et al.· The International Archives o...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
