Abstract

Hyperspectral imaging in the near infrared range (1000–1700 nm) was evaluated to identify different polyolefin flakes for quality assessment of recycled products. According to market requirements, the output of the recycling process of polyolefins must be high purity secondary polypropylene and polyethylene granulates. Hyperspectral images were acquired for selected plastic flakes coming from household waste, classified according to their typology. Spectra were analysed using principal component analysis to reduce the high dimensionality of data and for selecting effective wavelengths. Partial least square discriminant analysis was applied for classification of the two polyolefin typologies. Prediction results showed that it is possible to recognise the different plastic flakes with sensitivity and specificity ranging from 0.90 to 0.99. The developed procedures based on hyperspectral imaging can be utilised for quality control of the two pure polypropylene and polyethylene flow streams obtained by the innovative recycling process based on magnetic density separation.

© 2012 IM Publications LLP

Identification of overlapping plastic sheets using short-wavelength infrared hyperspectral imaging

Keisuke Ozawa
Opt. Express 31(8) 12328-12338 (2023)

Automated tea quality classification by hyperspectral imaging

Jiewen Zhao, Quansheng Chen, Jianrong Cai, and Qin Ouyang
Appl. Opt. 48(19) 3557-3564 (2009)

Spectral characterization and discrimination of synthetic fibers with near-infrared hyperspectral imaging system

Xiaoke Jin, Hafeezullah Memon, Wei Tian, Qinli Yin, Xiaofang Zhan, and Chengyan Zhu
Appl. Opt. 56(12) 3570-3576 (2017)

Focal-plane processing architectures for real-time hyperspectral image processing

Sek M. Chai, Antonio Gentile, Wilfredo E. Lugo-Beauchamp, Javier Fonseca, José L. Cruz-Rivera, and D. Scott Wills
Appl. Opt. 39(5) 835-849 (2000)

Blur-specific image quality assessment of microscopic hyperspectral images

Laura Quintana-Quintana, Samuel Ortega, Himar Fabelo, Francisco J. Balea-Fernández, and Gustavo M. Callico
Opt. Express 31(8) 12261-12279 (2023)