River Watch camera system
The Riverwatch Camera System uses deep-learning-based object detection to monitor and quantify floating plastic pollution in rivers in real time, enabling early interventions and providing valuable data for research and operational purposes.
Plastic pollution of rivers can appear unexpectedly as brought by floods and thus risking not only the natural values along rivers but infrastructures, too. Therefore, with the Riverwatch Camera System you can keep one eye always on the river and plan for upcoming interventions. The core of the system is the object detection tool, which is a deep-learning image-based artificial intelligence that the developers have trained specifically for this task. The camera system can detect and quantify floating plastic pollution in real time with low computational power and issue early alarms for intervention. The system using the detection model can help in measuring plastic loads in several river sections, thus providing essential data for a number of studies on plastic transport, and can also be used for operational purposes.
Related projects:
Future Plastik Ltd. developed the initial version of the detection model and camera system within the framework of the 2018-1.1.2-KFI-2018-00034 project funded by the Hungarian National Research, Development and Innovation Office.
http://www.futureplastik.hu/robo-collect
The Sustainable Development and Technologies National Programme of the
Hungarian Academy of Sciences (FFT NP FTA) supported the further development detection model and practical application at the Budapest University of Technology and Economics.
https://fenntarthatotechnologiak.hu/en/alprojekt/riverine-transport-of-macroplastics-and-their-potential-recycling/
References and publications:
Tikász G., Baranya S., Gyalai-Korpos, M. (2024) Folyami makroműanyag szennyezők detektálása képalapú mesterséges intelligenciával
Development partner:
Future Plastik Ltd
BME https://vit.bme.hu/node/1252?language=en