Near-infrared hyperspectral imaging for non-destructive classification of commercial tea products

Research output: Contribution to journalArticle

9 Citations (Scopus)

Abstract

Tea is the most consumed manufactured drink in the world. In recent years, various high end analytical techniques such as high-performance liquid chromatography have been used to analyse tea products. However, these techniques require complex sample preparation, are time consuming, expensive and require a skilled analyst to carry out the experiments. Therefore, to support rapid and non-destructive assessment of tea products, the use of near infrared (NIR) (950-1760 nm) hyperspectral imaging (HSI) for classification of six different commercial tea products (oolong, green, yellow, white, black and Pu-erh) is presented. To visualise the HSI data, linear (principal component analysis (PCA) and multidimensional scaling (MDS)) and non-linear (t-distributed stochastic neighbour embedding (t-SNE) and isometric mapping (ISOMAP)) data visualisation methods were compared. t-SNE provided separation of the six commercial tea products into three groups based on the extent of processing: minimally processed, oxidised and fermented. To perform the classification of different tea products, a multi-class error-correcting output code (ECOC) model containing support vector machine (SVM) binary learners was developed. The classification model was further used to predict classes for pixels in the HSI hypercube to obtain the classification maps. The SVM-ECOC model provided a classification accuracy of 97.41±0.16 % for the six commercial tea products. The methodology developed provides a means for rapid, non-destructive, in situ testing of tea products, which would be of considerable benefit for process monitoring, quality control, authenticity and adulteration detection.
LanguageEnglish
Pages70-77
Number of pages8
JournalJournal of Food Engineering
Volume238
Early online date13 Jun 2018
DOIs
Publication statusPublished - 1 Dec 2018

Fingerprint

Tea
tea
image analysis
process monitoring
adulterated products
Principal Component Analysis
Quality Control
quality control
analytical methods
principal component analysis
high performance liquid chromatography
methodology
High Pressure Liquid Chromatography

Keywords

  • imaging spectroscopy
  • hypercube
  • multivariate
  • data visualisation
  • neighbourhood methods

Cite this

@article{4b9e356cb240480e909835c3077fbc3e,
title = "Near-infrared hyperspectral imaging for non-destructive classification of commercial tea products",
abstract = "Tea is the most consumed manufactured drink in the world. In recent years, various high end analytical techniques such as high-performance liquid chromatography have been used to analyse tea products. However, these techniques require complex sample preparation, are time consuming, expensive and require a skilled analyst to carry out the experiments. Therefore, to support rapid and non-destructive assessment of tea products, the use of near infrared (NIR) (950-1760 nm) hyperspectral imaging (HSI) for classification of six different commercial tea products (oolong, green, yellow, white, black and Pu-erh) is presented. To visualise the HSI data, linear (principal component analysis (PCA) and multidimensional scaling (MDS)) and non-linear (t-distributed stochastic neighbour embedding (t-SNE) and isometric mapping (ISOMAP)) data visualisation methods were compared. t-SNE provided separation of the six commercial tea products into three groups based on the extent of processing: minimally processed, oxidised and fermented. To perform the classification of different tea products, a multi-class error-correcting output code (ECOC) model containing support vector machine (SVM) binary learners was developed. The classification model was further used to predict classes for pixels in the HSI hypercube to obtain the classification maps. The SVM-ECOC model provided a classification accuracy of 97.41±0.16 {\%} for the six commercial tea products. The methodology developed provides a means for rapid, non-destructive, in situ testing of tea products, which would be of considerable benefit for process monitoring, quality control, authenticity and adulteration detection.",
keywords = "imaging spectroscopy, hypercube, multivariate, data visualisation, neighbourhood methods",
author = "Puneet Mishra and Alison Nordon and Julius Tschannerl and Guoping Lian and Sally Redfern and Stephen Marshall",
year = "2018",
month = "12",
day = "1",
doi = "10.1016/j.jfoodeng.2018.06.015",
language = "English",
volume = "238",
pages = "70--77",
journal = "Journal of Food Engineering",
issn = "0260-8774",

}

TY - JOUR

T1 - Near-infrared hyperspectral imaging for non-destructive classification of commercial tea products

AU - Mishra, Puneet

AU - Nordon, Alison

AU - Tschannerl, Julius

AU - Lian, Guoping

AU - Redfern, Sally

AU - Marshall, Stephen

PY - 2018/12/1

Y1 - 2018/12/1

N2 - Tea is the most consumed manufactured drink in the world. In recent years, various high end analytical techniques such as high-performance liquid chromatography have been used to analyse tea products. However, these techniques require complex sample preparation, are time consuming, expensive and require a skilled analyst to carry out the experiments. Therefore, to support rapid and non-destructive assessment of tea products, the use of near infrared (NIR) (950-1760 nm) hyperspectral imaging (HSI) for classification of six different commercial tea products (oolong, green, yellow, white, black and Pu-erh) is presented. To visualise the HSI data, linear (principal component analysis (PCA) and multidimensional scaling (MDS)) and non-linear (t-distributed stochastic neighbour embedding (t-SNE) and isometric mapping (ISOMAP)) data visualisation methods were compared. t-SNE provided separation of the six commercial tea products into three groups based on the extent of processing: minimally processed, oxidised and fermented. To perform the classification of different tea products, a multi-class error-correcting output code (ECOC) model containing support vector machine (SVM) binary learners was developed. The classification model was further used to predict classes for pixels in the HSI hypercube to obtain the classification maps. The SVM-ECOC model provided a classification accuracy of 97.41±0.16 % for the six commercial tea products. The methodology developed provides a means for rapid, non-destructive, in situ testing of tea products, which would be of considerable benefit for process monitoring, quality control, authenticity and adulteration detection.

AB - Tea is the most consumed manufactured drink in the world. In recent years, various high end analytical techniques such as high-performance liquid chromatography have been used to analyse tea products. However, these techniques require complex sample preparation, are time consuming, expensive and require a skilled analyst to carry out the experiments. Therefore, to support rapid and non-destructive assessment of tea products, the use of near infrared (NIR) (950-1760 nm) hyperspectral imaging (HSI) for classification of six different commercial tea products (oolong, green, yellow, white, black and Pu-erh) is presented. To visualise the HSI data, linear (principal component analysis (PCA) and multidimensional scaling (MDS)) and non-linear (t-distributed stochastic neighbour embedding (t-SNE) and isometric mapping (ISOMAP)) data visualisation methods were compared. t-SNE provided separation of the six commercial tea products into three groups based on the extent of processing: minimally processed, oxidised and fermented. To perform the classification of different tea products, a multi-class error-correcting output code (ECOC) model containing support vector machine (SVM) binary learners was developed. The classification model was further used to predict classes for pixels in the HSI hypercube to obtain the classification maps. The SVM-ECOC model provided a classification accuracy of 97.41±0.16 % for the six commercial tea products. The methodology developed provides a means for rapid, non-destructive, in situ testing of tea products, which would be of considerable benefit for process monitoring, quality control, authenticity and adulteration detection.

KW - imaging spectroscopy

KW - hypercube

KW - multivariate

KW - data visualisation

KW - neighbourhood methods

UR - https://www.sciencedirect.com/journal/journal-of-food-engineering

U2 - 10.1016/j.jfoodeng.2018.06.015

DO - 10.1016/j.jfoodeng.2018.06.015

M3 - Article

VL - 238

SP - 70

EP - 77

JO - Journal of Food Engineering

T2 - Journal of Food Engineering

JF - Journal of Food Engineering

SN - 0260-8774

ER -