Autonomous Coastal Waste Detection and Localization Using UAV

Eyüp Koyun; Nur Hazal Çufalcı; Berke Akdoğan; Aytaç Uğur Yerden

doi:10.65582/aifsc.2026.010

Authors

Eyüp Koyun Istanbul Gedik University, Faculty of Engineering, Department of Computer Engineering, Istanbul, Türkiye. https://orcid.org/0009-0005-8571-6474
Nur Hazal Çufalcı Istanbul Gedik University, Faculty of Engineering, Department of Computer Engineering, Istanbul, Türkiye. https://orcid.org/0009-0001-4973-1256
Berke Akdoğan Istanbul Gedik University, Faculty of Engineering, Department of Computer Engineering, Istanbul, Türkiye. https://orcid.org/0009-0005-8916-235X
Aytaç Uğur Yerden Istanbul Gedik University, Faculty of Engineering, Department of Industrial Engineering, Istanbul, Türkiye. https://orcid.org/0000-0002-3886-802X

DOI:

https://doi.org/10.65582/aifsc.2026.010

Keywords:

Coastal Waste Detection, Offboard Processing, YOLOv12 Object Detection, Geospatial Mapping, Artificial Intelligence

Abstract

Waste that accumulates along coastlines—particularly in areas that are difficult for people to access—and is predominantly composed of plastic derivatives constitutes an environmental problem that threatens marine ecosystems, especially marine life, and concerns all countries, whether they have a coastline or not. This study aims to develop a ground-station-based, efficient waste detection and localization project using low-cost, easily accessible commercial unmanned aerial vehicles (UAVs). As the methodological approach, the attention-based and up-to-date deep learning architecture YOLOv12, which has been proven to be successful in object detection, was employed for detecting coastal waste. A robust training dataset was constructed using a hybrid dataset formed from coastal waste images captured over different surface types along the coastlines of Istanbul, together with open datasets. The main novelty of the system is the “VideoLog Synchronization Module,” which precisely matches the video stream with the flight data recorded by the drone. Through this module, each frame and the waste contained within it are labeled with geographic coordinates and made observable via an ASP.NET Core MVC–based interface. Results from tests conducted through drone flights indicate that the system detects small waste objects in complex backgrounds with high accuracy (mAP50: 0.783) and performs localization with a very high level of success. This study aims to enable municipalities, non-governmental organizations, as well as individuals and institutions involved in coastal waste detection and cleanup activities, to shift human effort from detection to collection, thereby saving labor and time.

References

Bochkovskiy, A., Wang, C. and Liao, H.M., 2020. YOLOv4: Optimal speed and accuracy of object detection. arXiv. DOI: https://doi.org/10.48550/arXiv.2004.10934.

Buleu, B., Robu, R. and Filip, I., 2025. A deep learning-based system for automatic license plate recognition using YOLOv12 and PaddleOCR. Applied Sciences, 15(14), 7833. DOI: https://doi.org/10.3390/app15147833.

DJI, 2025. DJI Mini 3 User Manual v1.2. SZ DJI Technology Co., Ltd., Shenzhen, China.

Duangsuwan, S. and Prapruetdee, P., 2024. Drone-enabled AI edge computing and 5G communication network for real- time coastal litter detection. Drones, 8(12), 750. DOI: https://doi.org/10.3390/drones8120750.

Eker, R., Alkan, E. and Aydın, A., 2021. A comparative analysis of UAV-RTK and UAV-PPK methods in mapping different surface types. European Journal of Forest Engineering, 7(1), 12–25. DOI: https://doi.org/10.33904/ejfe.938067.

Goodfellow, I., Bengio, Y. and Courville, A., 2016. Deep learning. Cambridge, MA: MIT Press.

Jiang, C., Zhou, X., Chen, H. and Liu, T., 2026. UAV positioning using GNSS: A review of the current status. Drones, 10(2), 91. DOI: https://doi.org/10.3390/drones10020091.

Kraft, M., Piechocki, M., Ptak, B. and Walas, K., 2021. Autonomous, onboard vision-based trash and litter detection in low altitude aerial images collected by an unmanned aerial vehicle. Remote Sensing, 13(5), 965. DOI: https://doi.org/10.3390/rs13050965.

Leira, F.S., Trnka, K., Fossen, T.I. and Johansen, T.A., 2015. A lightweight thermal camera payload with georeferencing capabilities for small fixed-wing UAVs. In: Proceedings of the 2015 International Conference on Unmanned Aircraft Systems (ICUAS), Denver, CO, USA, 9–12 June 2015, pp. 485–494. DOI: https://doi.org/10.1109/ICUAS.2015.7152327.

Leira, F.S., Helgesen, H.H., Johansen, T.A. and Fossen, T.I., 2021. Object detection, recognition, and tracking from UAVs using a thermal camera. Journal of Field Robotics, 38, 242–267. DOI: https://doi.org/10.1002/rob.21985.

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, Zurich, Switzerland, 6–12 September 2 014, pp. 740–755. DOI: https://doi.org/10.1007/978-3-319-10602-1_48.

Marin, D., Ciucă, A.-M., Filimon, A. and Stoica, E., 2024. Marine litter and plastic detection on Black Sea beaches by using unmanned aerial vehicle (UAV). Ovidius University Annals of Chemistry, 35(1), 43–50. DOI: https://doi.org/10.2478/auoc-2024-0006.

Merlino, S., Paterni, M., Berton, A. and Massetti, L., 2020. Unmanned aerial vehicles for debris survey in coastal areas: Long-term monitoring programme to study spatial and temporal accumulation of the dynamics of beached marine litter. Remote Sensing, 12(8), 1260. DOI: https://doi.org/10.3390/rs12081260.

Panwar, H., Gupta, P.K., Siddiqui, M.K., Morales-Menendez, R., Bhardwaj, P., Sharma, S. and Sarker, I.H., 2020. AquaVision: Automating the detection of waste in water bodies using deep transfer learning. Case Studies in Chemical and Environmental Engineering, 2, 100026. DOI: https://doi.org/10.1016/j.cscee.2020.100026.

Pfeiffer, R., Valentino, G., D’Amico, S., Piroddi, L., Galone, L., Calleja, S., Farrugia, R.A. and Colica, E., 2023. Use of UAVs and deep learning for beach litter monitoring. Electronics, 12(1), 198. DOI: https://doi.org/10.3390/electronics12010198.

Presidency of the Republic of Türkiye Digital Transformation Office, 2021. National artificial intelligence strategy (2021–2025). Ankara: Presidency of the Republic of Türkiye. Available at https://wp.oecd.ai/app/uploads/2021/12/Turkey_National_Artificial_Intelligence_Strategy_2021-2025.pdf [Accessed 11 May 2026].

Presidency of the Republic of Türkiye Strategy and Budget Directorate, 2024. 12th development plan (2024–2028). Ankara: Presidency of the Republic of Türkiye. Available at https://www.sbb.gov.tr/wp- content/uploads/2025/03/Twelfth-Development-Plan_2024-2028.pdf [Accessed 11 May 2026].

Republic of Türkiye Ministry of Industry and Technology, 2023. 2030 industry and technology strategy. Ankara: Republic of Türkiye Ministry of Industry and Technology. Available at https://www.sanayi.gov.tr/assets/pdf/plan- program/2030IndustryAndTechnologyStrategyENG.pdf [Accessed 11 May 2026].

Terzi, Y., Gündoğdu, S., Yandi, İ., Altinpinar, İ., Öztürk, R.Ç. and Gedik, K., 2025. Marine litter on the Turkish Black Sea shoreline: Abundance, composition, and sources. Waste Management, 205, 115027. DOI: https://doi.org/10.1016/j.wasman.2025.115027.

Tian, Y., Ye, Q. and Doermann, D.S., 2025. YOLOv12: Attention-centric real-time object detectors. arXiv. DOI: https://doi.org/10.48550/arXiv.2502.12524.

Verma, R., Taneja, H., Singh, K.D. and Singh, P.D., 2024. Enhancing data analytics in environmental sensing through cloud IoT integration. Journal of Climate Change, 10(2), 6. DOI: https://doi.org/10.3233/JCC240013.

Yerden, A.U., Şenol, S., Kara, M. and Dilibal, S., 2025. xT-STRIDE threat model for unmanned air vehicle security. International Journal of Information Security, 24, 169. DOI: https://doi.org/10.1007/s10207-025-01082-4.

Zhu, H., Zhu, D., Qin, X. and Guo, F., 2023. Efficient and accurate beach litter detection method based on QSB-YOLO. International Journal of Advanced Computer Science and Applications, 14(7). DOI: https://doi.org/10.14569/IJACSA.2023.0140735.