Key points are not available for this paper at this time.
Due to the geology of the Mediterranean coastline zone and insufficient waste management in many nations, the Mediterranean Sea has become overflowed with plastic litter attributed to its dense population and high level of tourism activity. To mitigate the plastic pollution, protect marine life, and preserve the ecological balance a series of novel approaches for monitoring and detection of marine litter in the Mediterranean sea are needed. The primary objective of this study is to demonstrate the feasibility of using AI-driven aerial drones for the detection of plastic hotspots on beaches, followed by the use of a monitoring app for community-led plastic pollution monitoring and cleanup initiatives that were held at Saidia beach in Morocco in November 2023. For that purpose, artificial intelligence was tested to quantify and identify litter on beaches using drones that flew over the beach being monitored. Specifically, the drone's video stream is processed by an algorithm that first segments (in polygons) the objects in the video stream and then through deep learning (DL) each object is identified to categorise it as plastic or general waste. The acquired images are then used to train the DL algorithm in order to constantly improve the recognition performance of plastic and other generic waste types. This technique will allow the observation in detail of the monitoring area before and after the monitoring/clean up event, and thus, it can serve as a method to validate the grade of execution of the activity and analysis of the monitored/cleaned area. The focus on citizen science is essential to connect the public with the technologies that will allow them to collaborate in the collection of methodical data that can complement the existing data for a more detailed analysis.Together with the drones, another approach is the new app that will include the option to collect data for beach monitoring and for beach clean-ups. Created to function in both IOS and Android operating systems, this smart app for collecting marine litter monitoring data features an intuitive user interface and other advanced tools to enable even non-professional users to properly collect scientific data. The app also is designed to be used simultaneously by multiple users, that is, to collect data from multiple devices and referring to a single monitoring event. At the conclusion of the event, all collected data can be easily reviewed and supplemented with other advanced metadata for subsequent analysis and sharing activities, as well as then shared in the European repository of the EMODnet ML. The compilation of data from these techniques, to be tested on different demo sites, together with the results of future replications in other areas and the input of data from citizens and external organisations, will be the next step to facilitate a more holistic approach to tackle the crucial situation the Mediterranean sea is facing nowadays due the uncontroled discharge of plastics in its waters. Acknowledgements The authors acknowledge financial support from the European Unions HORIZON EUROPE innovation program for the project REMEDIES awarded under Grant Agreement No. 101093964.
Delgado et al. (Fri,) studied this question.