Rapid and Nondestructive On-Site Classification Method for Consumer-Grade Plastics Based on Portable NIR Spectrometer and Machine Learning

Abstract

The classification of plastic waste before recycling is of great significance to achieve effective recycling. In order to achieve rapid, nondestructive, and on-site detection, a portable near-infrared spectrometer was used in this study to obtain the diffuse reflectance spectrum for both standard and commercial plastics made by ABS, PC, PE, PET, PP, PS, and PVC. After applying a series of pretreatments, the principal component analysis (PCA) was used to analyze the cluster trend. K-nearest neighbor (KNN), support vector machine (SVM), and back propagation neural network (BPNN) classification models were developed and evaluated, respectively. The result showed that different plastics could be well separated in top three principal components space after pretreatment, and the classification models performed excellent classification results and high generalization capability. This study indicated that the portable NIR spectrometer, integrated with chemometrics, could achieve excellent performance and has great potential in the field of commercial plastic identification.

Qualitative Analysis of Lambda-Cyhalothrin on Chinese Cabbage Using Mid-Infrared Spectroscopy Combined with Fuzzy Feature Extraction Algorithms

“…Even though some of the traditional methods might be more sensitive [7][8][9][10][11], they require multistep sample preparation, are time-consuming and can only be conducted in the laboratory [12]. The spectroscopy method that we used could realize on-site detection of pesticide residues with portable spectrometers [31]. Moreover, there is no requirement for the use of chemical reagents for detection, which reduces the cost and effect on the environment.…”

Deep Learning for Chemometric Analysis of Plastic Spectral Data from Infrared and Raman Databases

“…PCA was used in earlier chemometrics work to identify the potential on distinguishing plastics by IR spectra (De Biasio et al, 2010) and LIBS spectra (Grégoire et al, 2011). It also continues to be used as a pre-processing technique prior to machine learning model training (Musu et al, 2019;Yan et al, 2021;Yang et al, 2020;Zhu et al, 2019) Supervised machine learning techniques include PLS, kNN and SVM are also popular chemometric techniques. For PLS, latent variables are constructed that best explain the relationship between the spectral data and the output label.…”

“…Liu et al, 2019;Pieszczek & Daszykowski, 2019;Saeki et al, 2003;Sato et al, 2002). The kNN algorithm classifies data based on the majority class of its surrounding neighbors (Costa et al, 2017;Yang et al, 2020). SVM algorithm works based on the construction of a hyperplane that serves as a decision boundary between different classes (Musu et al, 2019;Yang et al, 2020;Yu et al, 2014;Zhu et al, 2019).…”

“…The kNN algorithm classifies data based on the majority class of its surrounding neighbors (Costa et al, 2017;Yang et al, 2020). SVM algorithm works based on the construction of a hyperplane that serves as a decision boundary between different classes (Musu et al, 2019;Yang et al, 2020;Yu et al, 2014;Zhu et al, 2019). These machine learning techniques have been found to give good results with IR spectra (Bonifazi et al, 2014;Calvini et al, 2018;Karaca et al, 2013;Rani et al, 2019;Said et al, 2020;S Serranti et al, 2020;Silvia Serranti et al, 2012Ulrici et al, 2013;Wu et al, 2020;Yang et al, 2020;Zhu et al, 2019), Raman spectra (Allen et al, 1999;L.…”

“…SVM algorithm works based on the construction of a hyperplane that serves as a decision boundary between different classes (Musu et al, 2019;Yang et al, 2020;Yu et al, 2014;Zhu et al, 2019). These machine learning techniques have been found to give good results with IR spectra (Bonifazi et al, 2014;Calvini et al, 2018;Karaca et al, 2013;Rani et al, 2019;Said et al, 2020;S Serranti et al, 2020;Silvia Serranti et al, 2012Ulrici et al, 2013;Wu et al, 2020;Yang et al, 2020;Zhu et al, 2019), Raman spectra (Allen et al, 1999;L. Chen et al, 2017;Musu et al, 2019;Silva & Wiebeck, 2019) and LIBS spectra (Costa et al, 2017;Vahid Dastjerdi et al, 2018;Yan et al, 2021;Yu et al, 2014).…”