REMOTE MICROPLASTIC DETECTOR FOR WATER BODIES

Abstract

Microplastic pollution in aquatic environments has emerged as a critical environmental challenge due to its adverse impacts on ecosystems and human health. Conventional detection methods are labor-intensive, time-consuming, and limited in spatial coverage, necessitating the development of efficient and scalable solutions. This study presents an integrated approach for detecting and mapping microplastics using remote sensing, deep learning, and hyperspectral image analysis. The system employs geo-referenced RGB images processed through a YOLOv8-based deep learning model to identify plastic waste in water bodies. Additionally, hyperspectral data is utilized to generate classification and abundance maps using spectral unmixing and Maximum Abundance Classification techniques. The proposed framework also detects large floating debris using advanced segmentation models, enabling comprehensive monitoring of surface-level pollution. Embedded systems such as Raspberry Pi integrated with camera modules support real-time data acquisition and deployment. Experimental results demonstrate improved detection accuracy and efficient mapping capabilities compared to traditional approaches. This integrated methodology provides a scalable, costeffective, and automated solution for environmental monitoring. The findings contribute to better understanding of microplastic distribution and support policy-making and sustainable management strategies for mitigating plastic pollution in aquatic ecosystems.