Principal Investigator: Constantinos Hadjistassou
Affiliation: School of Sciences and Engineering, University of Nicosia.
Aims and Objectives
- Marine debris is a problem of global dimensions. Adverse impacts from marine debris range from fatal incidents of fish and marine birds to ghost-fishing and the introduction of alien species and bio-invasion. The durability, ability to float, toxic nature, sheer volume and ubiquity of marine litter calls for urgent action if we are to avert a major ecological catastrophe.
- This research project considered three ways of measuring and geographically mapping the presence of plastic debris in the Mediterranean Sea. The first focused on building a prototype device to be installed onboard ships able to detect and recognise in-situ plastics using object recognition techniques. The second system concerned fitting plastic detection sensors at the ship’s water ballast intake for detecting smaller size plastic debris.
- To-date most of the approaches used to measure and map the distribution of plastic debris utilize either surface net tows or visual survey transects which require expensive and prolonged survey expeditions, are prone to human error and are usually limited in scope. Herein was studied a third novel method for tracing, identifying and quantifying macroscopic and microscopic marine debris.
Outcomes - Summary
Addressing the problem of plastic debris in the world’s oceans is one of the most challenging environmental problems of our time. To overcome the existing, time consuming and of limited effectiveness manual methods for detecting macro- and micro-plastics this research introduced the use of novel unconventional methods for tracing and classifying marine plastics. For macro-plastics, the development of a deep learning method capable of detecting and identifying floating macro-plastics was shown to yield a high classification accuracy. In parallel, for the detection of micro-plastics a hydro-acoustic device was used with promising early results.