Scholarlink

<p>Plastic pollution in riverine and coastal environments has emerged as a critical environmental concern, threatening marine ecosystems, biodiversity, and public health. This study presents an automated and scalable machine learning-based approach for identifying and monitoring river plastic litter in coastal regions using Sentinel-2 multispectral satellite imagery. By integrating image preprocessing, spectral feature extraction, and supervised classification, the proposed method aims to detect plastic debris in dynamic aquatic environments accurately. Five machine learning algorithms Decision Tree (DT), Naïve Bayes (NB), Support Vector Classifier (SVC), Random Forest (RF), and Artificial Neural Network (ANN) were applied and evaluated using accuracy, precision, and sensitivity metrics. The SVC algorithm demonstrated the best overall performance with an accuracy of 0.99, precision of 0.95, and sensitivity of 0.93. ANN followed closely with 0.98 accuracy, 0.96 precision, and 0.94 sensitivity, while RF achieved 0.94 accuracy, 0.89 precision, and 0.88 sensitivity. DT and NB recorded lower accuracies of 0.85 and 0.88 respectively. These results highlight the high potential of SVC and ANN for precise classification of plastic waste in coastal waters. The study provides a cost-effective, replicable solution that supports frequent environmental monitoring and marine pollution management. Environmental agencies and policymakers can utilize the proposed methodology for targeted cleanup efforts, regulatory planning, and sustainable marine conservation. This research advances the use of Earth observation and artificial intelligence for addressing one of the most pressing challenges in coastal ecosystems.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Machine Learning Based Identification of River Plastic Litter in Coastal Region Using Sentinel-2 Data: An Automated Approach for Marine Pollution Monitoring and Management

  • 期刊论文

  • 发表日期：

    2025年10月20日

查看全文
  • M. Vadivel,
  • Nada Alzaben,
  • Malak Zayed Alamri,
  • Menwa Alshammerid,
  • S. V. N. Pammi,
  • V. Priya

摘要

Plastic pollution in riverine and coastal environments has emerged as a critical environmental concern, threatening marine ecosystems, biodiversity, and public health. This study presents an automated and scalable machine learning-based approach for identifying and monitoring river plastic litter in coastal regions using Sentinel-2 multispectral satellite imagery. By integrating image preprocessing, spectral feature extraction, and supervised classification, the proposed method aims to detect plastic debris in dynamic aquatic environments accurately. Five machine learning algorithms Decision Tree (DT), Naïve Bayes (NB), Support Vector Classifier (SVC), Random Forest (RF), and Artificial Neural Network (ANN) were applied and evaluated using accuracy, precision, and sensitivity metrics. The SVC algorithm demonstrated the best overall performance with an accuracy of 0.99, precision of 0.95, and sensitivity of 0.93. ANN followed closely with 0.98 accuracy, 0.96 precision, and 0.94 sensitivity, while RF achieved 0.94 accuracy, 0.89 precision, and 0.88 sensitivity. DT and NB recorded lower accuracies of 0.85 and 0.88 respectively. These results highlight the high potential of SVC and ANN for precise classification of plastic waste in coastal waters. The study provides a cost-effective, replicable solution that supports frequent environmental monitoring and marine pollution management. Environmental agencies and policymakers can utilize the proposed methodology for targeted cleanup efforts, regulatory planning, and sustainable marine conservation. This research advances the use of Earth observation and artificial intelligence for addressing one of the most pressing challenges in coastal ecosystems.