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.

From visual estimates to fully automated sensor-based measurements of plant disease severity: status and challenges for improving accuracy

“…Thus, airborne or space borne systems have lower spatial resolution compared to near-range systems. Data preprocessing and analysis is closely linked and individually designed depending on the sensor, setup and purpose of measuring (Behmann et al 2015a;Mishra et al 2018).…”

Grade Identification of Tieguanyin Tea Using Fluorescence Hyperspectra and Different Statistical Algorithms

“…HSI is an advanced analytical technique that combines traditional imaging and spectroscopy techniques to simultaneously acquire spatial and spectral information (Dale, Thewis, & Boudry, 2013). By combining appropriate analytical methods, it cannot only detect the main components of tea (Zhao, Wang, Ouyang, & Chen, 2011), but also effectively classify tea varieties (Puneet et al, 2018;Sun et al, 2018). Fluorescence spectroscopy, which is a mature, simple, fast, accurate and nondestructive technique, has been found to be highly effective in tea classification and quality assessment due to its high sensitivity (Dong et al, 2014;Seetohul, Scott, O'Hare, Ali, & Islam, 2013).…”

Total phenolic content prediction inFlos Loniceraeusing hyperspectral imaging combined with wavelengths selection methods

“…During the image acquisition, the distance between the platform and the lens was 250 mm, the moving speed of the platform was 1.2 mm/s, and the exposure time of the spectral camera was 90 ms. A total of 350 hyperspectral images were obtained. The acquired original hyperspectral images should be corrected by white and black reference images for reducing the impacts of the dark current noise and uneven illumination (Li et al, ; Liu, Wang, et al, ; Mishra et al, ).…”