2025-09-03 ノースカロライナ州立大学（NC State）
＜関連情報＞
- https://news.ncsu.edu/2025/09/how-hyper-detailed-cameras-will-make-recycling-more-efficient/
- https://www.sciencedirect.com/science/article/abs/pii/S2590238525004084#preview-section-references
エンドメンバ抽出と存在量検出を用いた廃棄物材料のリアルタイム特性評価および回収のためのハイパースペクトルイメージング Hyperspectral imaging for real-time waste materials characterization and recovery using endmember extraction and abundance detection
Mariangeles Salas, Simran Singh, Raman Rao, Raghul Thiyagarajan, Ashutosh Mittal, John Yarbrough, Anand Singh, Lucian Lucia, Lokendra Pal
Matter Available online: 1 August 2025
DOI:https://doi.org/10.1016/j.matt.2025.102365
Graphical abstract
Highlights
- Hyperspectral unmixing enables precise material identification in complex objects
- PPI and SMACC methods show differences in spectral signature extraction of materials
- FCLS accurately quantifies material abundance in complex multi-component objects
- Enables real-time characterization of materials for energy recovery and recycling
Progress and potential
Every person on Earth is affected by less than optimal waste management. Municipal solid waste (MSW) generation and accumulation are rising at an alarming rate due to rapid population growth and urbanization, posing significant risks to human health, public safety, and ecosystems. According to the US Environmental Protection Agency, approximately 300 million tons of MSW were generated in 2018, of which only 32% was recycled and 50% was landfilled. This represents a significant loss of valuable resources and creates both environmental and economic challenges.
Our study addresses this challenge by demonstrating the use of hyperspectral imaging (HSI) and unmixing techniques to accurately identify material types in complex, everyday items such as disposable coffee cups. HSI goes beyond what the human eye can see by capturing pixel-level data across a wide spectrum of wavelengths. It determines what materials are made of based on how they reflect light, even when they appear visually identical. The result is a robust spectroscopic method for accurate real-time characterization of complex multi-material objects, enabling smart manufacturing, efficient recycling, energy recovery, and materials circularity for sustainable waste management and industrial applications.
Summary
Hyperspectral imaging, combined with advanced spectral unmixing techniques and artificial intelligence, offers a powerful solution for improving material identification and classification. This study evaluates the effectiveness of the pixel purity index and the sequential maximum angle convex cone algorithms in extracting and validating spectral signatures from pure samples of paper components (cellulose and lignin) and plastic (polypropylene). Principal-component analysis showed that both algorithms captured nearly all relevant variance for the tested materials. Spectral signatures were compared using the spectral angle mapper, revealing high similarity in the short-wave infrared region and greater variability in the visible near-infrared range. The methodology was then applied to a disposable coffee cup to detect and quantify mixed materials, accurately estimating material abundance and object area with less than 1% error. This approach enhances material classification, supporting product verification, quality control, and automated sorting for sustainable waste management and resource recovery.
