School of Science, Royal Melbourne Institute of Technology University, 124 La Trobe St, Melbourne 3000, Victoria, Australia; ITC, University of Twente, Hengelosestraat 99, Enschede 7514 AE, Netherlands.
School of Science, Royal Melbourne Institute of Technology University, 124 La Trobe St, Melbourne 3000, Victoria, Australia.
Mar Pollut Bull. 2024 Dec;209(Pt A):117124. doi: 10.1016/j.marpolbul.2024.117124. Epub 2024 Oct 22.
Plastic pollution on shorelines poses a significant threat to coastal ecosystems, underscoring the urgent need for scalable detection methods to facilitate debris removal. In this study, the Beached Plastic Debris Index (BPDI) was developed to detect plastic accumulation on beaches using shortwave infrared spectral features. To validate the BPDI, plastic targets with varying sub-pixel covers were placed on a sand spit and captured using WorldView-3 satellite imagery. The performance of the BPDI was analysed in comparison with the Normalized Difference Plastic Index (NDPI), the Plastic Index (PI), and two hydrocarbon indices (HI, HC). The BPDI successfully detected the plastic targets from sand, water, and vegetation, outperforming the other indices and identifying pixels with <30 % plastic cover. The robustness of the BPDI suggests its potential as an effective tool for mapping plastic debris accumulations along coastlines.
海滩塑料污染对沿海生态系统构成重大威胁,这突显了需要可扩展的检测方法来促进碎片清除。在这项研究中,开发了海滩塑料碎片指数(BPDI),以便使用短波红外光谱特征来检测海滩上的塑料堆积。为了验证 BPDI,在沙堤上放置了具有不同亚像元覆盖的塑料目标,并使用 WorldView-3 卫星图像进行了捕获。分析了 BPDI 与归一化差异塑料指数(NDPI)、塑料指数(PI)以及两种碳氢化合物指数(HI、HC)的性能。BPDI 成功地从沙、水和植被中检测到塑料目标,表现优于其他指数,并识别出塑料覆盖面积<30%的像素。BPDI 的稳健性表明其有可能成为沿海岸线绘制塑料碎片堆积地图的有效工具。