15:30 〜 17:00
[MIS17-P01] Progress in Identification and Characterizing Plastic Marine Debris in Ocean using Spectral Information
★Invited Papers
キーワード：spectral properties, marine debris, normalize difference spectral index, ocean plastic
Worldwide production of plastic reaches 367 million tons annually, with a significant portion entering the marine environment due to intensified human activities. Hence, plastic marine debris has become a crucial global environmental concern that demands immediate attention. In the meantime, visual inspections, or calibrated equipment for in situ measurements are time-consuming and require a considerable workforce. Meanwhile, remote sensing, through the air or space-borne spectral imaging, requires a thorough understanding of the spectral properties of plastics in the ocean. Determining plastic’s spectral properties in a marine environment is a significant challenge because of several factors, including camera and light angle, water conditions, plastic debris characteristics, and non-plastic materials.
Here, the study presents preliminary laboratory spectral measurements performed on 120 samples, including commercial-off-the-shelf (COTS) plastics, non-plastics, and marine debris. The debris was gathered from Nokobiura Bay, Goto Islands, Nagasaki, Japan. The measurement was done using an InGaAs-LCTF-type hyperspectral imaging camera with a wavelength range of 400-1600 nm and a fixed 10 nm resolution. In addition, different sample conditions were included in the measurement to simulate the sample's lifetime experiences. The spectral data was then assigned to a spectral library.
Subsequently, a normalized difference spectral index (NDSI) with a discriminant analysis (DA) algorithm was used to identify the plastic by calculating the highest discrimination accuracy at a specific wavelength combination. The method was compared with common DA algorithms to assess its efficacy. Future work will incorporate polarization and bidirectional reflectance distribution function (BRDF) using the NDSI information, enabling rapid identification and simplifying camera system complexity for future drone or satellite installations. This research is supported by Dr. Shin’ichiro Kako of Kagoshima University, Dr. Daisuke Matsuoka of the Japan Agency for Marine-Earth Science and Technology (JAMSTEC), and Mr. Shinya Kimura.
Here, the study presents preliminary laboratory spectral measurements performed on 120 samples, including commercial-off-the-shelf (COTS) plastics, non-plastics, and marine debris. The debris was gathered from Nokobiura Bay, Goto Islands, Nagasaki, Japan. The measurement was done using an InGaAs-LCTF-type hyperspectral imaging camera with a wavelength range of 400-1600 nm and a fixed 10 nm resolution. In addition, different sample conditions were included in the measurement to simulate the sample's lifetime experiences. The spectral data was then assigned to a spectral library.
Subsequently, a normalized difference spectral index (NDSI) with a discriminant analysis (DA) algorithm was used to identify the plastic by calculating the highest discrimination accuracy at a specific wavelength combination. The method was compared with common DA algorithms to assess its efficacy. Future work will incorporate polarization and bidirectional reflectance distribution function (BRDF) using the NDSI information, enabling rapid identification and simplifying camera system complexity for future drone or satellite installations. This research is supported by Dr. Shin’ichiro Kako of Kagoshima University, Dr. Daisuke Matsuoka of the Japan Agency for Marine-Earth Science and Technology (JAMSTEC), and Mr. Shinya Kimura.