Applying the YOLOv11 Model to Detect and Segment Shrimp Farms in Georeferenced Images

• Alisson C. Ferreira IFCE
• Luís J. R. Oliveira UERN
• Ana Carla C. da Silva IFCE
• Bruno S. Ursulino IFCE
• Davidson A. Nunes IFCE
• Danielo G. Gomes UFC
• Raimundo V. C. F. IFCE

DOI: https://doi.org/10.5753/wcama.2025.7190

Abstract

The monitoring of shrimp farms is essential for tracking aquaculture production and identifying new cultivation areas. This study proposes a machine learning-based approach using the YOLOv11 architecture to detect and monitor farms in the states of Ceará and Rio Grande do Norte. A dataset of about 200 georeferenced images from the Copernicus© platform was created, covering the Red (R), Green (G), Blue (B), and Near-infrared (NIR) spectra, with annotations made in Roboflow©. The model was trained using the hold-out method (80% training, 20% validation) and evaluated using precision, recall, mAP50, and mAP50-95 metrics. The results confirm the effectiveness of the approach in identifying shrimp farms and detecting new cultivation areas, contributing to aquaculture monitoring and policy development.

References

Dong, D., Shi, Q., Hao, P., Huang, H., Yang, J., Guo, B., and Gao, Q. (2024). Intelligent detection of marine offshore aquaculture with high-resolution optical remote sensing images. Journal of Marine Science and Engineering, 12(1012).

Gomes, J. V. E. (2022). Detecção de objetos com a arquitetura yolo. Master’s thesis, Universidade Federal de Ouro Preto, João Monlevade, MG.

Greenstreet, W., Bucini, G., Bowen, M., and Radeloff, V. C. (2023). Detecting aquaculture with deep learning in a low-data setting. Remote Sensing in Ecology and Conservation.

IBGE (2023). Valor da produção da pecuária e aquicultura chega a R$ 122,4 bilhões em 2023. Acesso em 29 de dezembro de 2024.

Igarashi, M. A. (2022). Sinopse sobre os aspectos do potencial da produção e prospectos para o desenvolvimento do cultivo de camarão marinho no brasil, especialmente no nordeste brasileiro. Revista Semiárido De Visu, 10(3).

Jocher, G. and Qiu, J. (2024). Ultralytics yolo11.

Lin, T.-Y., Maire, M., Belongie, S., Hays, J., Perona, P., Ramanan, D., Dollár, P., and Zitnick, C. L. (2014). Microsoft coco: Common objects in context. In Computer vision–ECCV 2014: 13th European conference, zurich, Switzerland, September 6-12, 2014, proceedings, part v 13, pages 740–755. Springer.

Quaade, S., Vallebueno, A., Alcabes, O. D. N., Rodolfa, K. T., and Ho, D. E. (2024). Remote sensing and computer vision for marine aquaculture. Science Advances, 10(eadn4944).

Redmon, J., Divvala, S., Girshick, R., and Farhadi, A. (2016). You only look once: Unified, real-time object detection.

Werneck, P. R., Macedo, H. R., Morsoleto, F. M. d. S., Lira, K. C. d. S., and Feiden, A. (2023). Use of georeferencing as a sustainable planning tool for aquaculture production in small municipalities: a case study for ouro verde do oeste/pr. Research, Society and Development, 12(3):e17012338683.