Search

Ash is a testing index with both health inspection value and quality decision value, and it is an essential detection item in the import and export trade of tea. To realize the rapid and effective quantitative analysis of ash content in tea, this study proposed the use of a homemade miniature near-infrared (NIR) spectroscopy combined with multivariate analysis for the rapid detection of ash content in black tea. First, NIR data of black tea samples from different countries were acquired and optimized by the spectral preprocessing method. Then, the optimized pre-processed spectral data were used as features, and four feature wavelength selection algorithms, such as competitive adaptive reweighted sampling, iteratively retaining informative variables (IRIV), variable combination population analysis (VCPA)-IRIV, and interval variable iterative space shrinkage approach (IVISSA), were utilized to optimize the feature spectra. Finally, the support vector machine regression (SVR) algorithm was employed to construct the quantitative models of ash content in black tea by combining the optimal wavelengths obtained from the four feature selection methods mentioned above. The experimental results showed that the IVISSA-SVR model had the best performance, with correlation coefficient (R p ), root mean square errors of prediction (RMSEP), and relative prediction deviation (RPD) of 0.9546, 0.0192, and 5.59 for the prediction set, respectively. The results demonstrate that a miniature NIR sensing system combined with chemometrics as an effective analytical tool can realize the rapid detection of ash content in black tea., Competing Interests: Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2023 Elsevier B.V. All rights reserved.)

The taste of tea is one of the key indicators in the evaluation of its quality and is a key factor in its grading and market pricing. To objectively and digitally evaluate the taste quality of tea leaves, miniature near-infrared (NIR) spectroscopy and electronic tongue (ET) sensors are considered effective sensor signals for the characterization of the taste quality of tea leaves. This study used micro-NIR spectroscopy and ET sensors in combination with data fusion strategies and chemometric tools for the taste quality assessment and prediction of multiple grades of black tea. Using NIR features and ET sensor signals as fused information, the data optimization based on grey wolf optimization, ant colony optimization (ACO), particle swarm optimization, and non-dominated sorting genetic algorithm II were employed as modeling features, combined with support vector machine (SVM), extreme learning machine and K -nearest neighbor algorithm to build the classification models. The results obtained showed that the ACO-SVM model had the highest classification accuracy with a discriminant rate of 93.56%. The overall results reveal that it is feasible to qualitatively distinguish black tea grades and categories by NIR spectroscopy and ET techniques.

Background: Tea (Camellia sinensis L) is a highly nutritious beverage with commercial value globally. However, it is at risk of economic fraud. This study aims to develop a powerful evaluation method to distinguish Chinese official Dianhong tea from various other categories, employing hyperspectral imaging (HSI) technology and chemometric algorithms., Results: Two matrix statistical algorithms encompassing a gray-level co-occurrence matrix (GLCM) and a gradient co-occurrence matrix (GLGCM) are used to extract HSI texture data. Three novel spectral variable screening methods are utilized to select wavenumbers of near-infrared (NIR) spectra: iteratively retaining informative variables (IRIV), interval random frog, and variable combination population analysis. Feature fusion of image texture characteristics and spectra data are the eigenvectors for model building. Authentic classification models are constructed using the extreme learning machine approach and the least squares support vector machine (LSSVM) approach, coupling them with features from wavelength extraction techniques for assessing the quality of Dianhong black tea. The results demonstrate that the LSSVM model using fused data (IRIV + GLGCM) provides the best results and achieves a predictive precision of 99.57%., Conclusion: This study confirms that HSI coupled with LSSVM is effective in differentiating authentic Dianhong black tea samples. © 2020 Society of Chemical Industry., (© 2020 Society of Chemical Industry.)

The main objectives of the study are to understand and explore critical feature wavelengths of the obtained near-infrared (NIR) data relating to dianhong black tea quality categories, we propose a multi-variable selection strategy based on the variable space optimization from big to small which is the kernel idea of a variable combination of the improved genetic algorithm (IGA) and particle swarm optimization (PSO) in this study. A rapid description based on the NIR technology is implemented to assess black tea tenderness and rankings. First, 700 standard samples from dianhong black tea of seven quality classes are scanned using a NIR system. The raw spectra acquired are preprocessed by Savitzky-Golay (SG) filtering coupled with standard normal variate transformation (SNV). Then, the multi-variable selection algorithm (IGA-PSO) is applied to compare with the single method (the IGA and PSO) and search the optimal characteristic wavelengths. Finally, the identification models are developed using a decision tree (DT), partial least-squares discriminant analysis (PLS-DA), and support vector machine (SVM) based on different kernel functions combined with the effective features from the above variables screening paths for the discrimination of black tea quality. The results show that the IGA-PSO-SVM model with a radial basis function achieves the best predictive results with the correct discriminant rate (CDR) of 95.28% based on selected four characteristic variables in the prediction process. The overall results demonstrate that NIR combined with a multi-variable selection method can constitute a potential tool to understand the most important features involved in the evaluation of dianhong black tea quality helping the instrument manufacturers to achieve the development of low-cost and handheld NIR sensors., Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2020 Elsevier B.V. All rights reserved.)

Caffeine and catechin are two main components of instant green tea, and are essential components of tea quality. This paper mainly focuses on the feasibility of rapidly determining instant green tea components by using a portable near infrared (NIR) spectrometer. The two main components (caffeine and catechin) were studied. In addition, the instrument performance levels of portable and benchtop NIR spectrometers were studied and compared. Quantitative models developed using portable and benchtop spectrometers for measuring caffeine, total catechins, and four individual catechins were established and compared. After preprocessing using standard normal variate (SNV), the Rp values of the caffeine, total catechins, (-)-epigallocatechin, (-)-epigallocatechin 3-gallate, (-)-epicatechin, and (-)-epicatechin gallate in the partial least squares models for a portable NIR spectrometer were 0.974, 0.962, 0.669, 0.945, 0.942 and 0.905, respectively. For a benchtop NIR spectrometer, Rp values were 0.993, 0.958, 0.883, 0.955, 0.966 and 0.936, respectively. Passing-Bablok regression method results indicated no significant differences between the two instruments. A genetic algorithm (GA) and the successive projections algorithm (SPA) were used to screen the wavelength of the NIR spectrum and establish the model. The GA obtained more robust modeling results. This study concludes that the developed portable spectroscopy system combined with appropriate variable selection methods can be effectively used for rapid determination of caffeine, total catechins, and four individual catechins in instant green tea., Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2020 Elsevier B.V. All rights reserved.)

The evaluation of tea quality tended to be subjective and empirical by human panel tests currently. A convenient analytical approach without human involvement was developed for the quality assessment of tea with great significance. In this study, near-infrared hyperspectral imaging (HSI) combined with multiple decision tree methods was utilized as an objective analysis tool for delineating black tea quality and rank. Data fusion that integrated texture features based on gray-level co-occurrence matrix (GLCM) and short-wave near-infrared spectral features were as the target characteristic information for modeling. Three different types of supervised decision tree algorithms (fine tree, medium tree, and coarse tree) were proposed for the comparison of the modeling effect. The results indicated that the performance of models was enhanced by the multiple perception feature fusion. The fine tree model based on data fusion obtained the best predictive performance, and the correct classification rate (CCR) of evaluating black tea quality was 93.13% in the prediction process. This work demonstrated that HSI coupled with intelligence algorithms as a rapid and effective strategy could be successfully applied to accurately identify the rank quality of black tea., Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2020 Elsevier B.V. All rights reserved.)

From the perspective of combating fraud issues and examining keemun black tea properties, there was a contemporary urgent demand for a keemun black tea rankings identification system. Current rapid evaluation systems had been mainly developed for green tea grade evaluation, but there was space for improvement to establish a highly robust model. The present study proposed cognitive spectroscopy that combined near infrared spectroscopy (NIRS) with multivariate calibration and feature variable selection methods. We defined "cognitive spectroscopy" as a protocol that selects characteristic information from complex spectral data and showed optimal results without human intervention. 700 samples representing keemun black tea from seven quality levels were scanned applying an NIR sensor. To differentiate which wavelength variables of the acquired NIRS data carry key and feature information regarding keemun black tea grades, there were four different variables screening approaches, namely genetic algorithm (GA), successive projections algorithm (SPA), competitive adaptive reweighted sampling (CARS), and shuffled frog leaping algorithm (SFLA), were compared in this study. Cognitive models were developed using least squares support vector machine (LSSVM), back propagation neural network (BPNN) and random forest (RF) methods combined with the optimized characteristic variables from the above variables selection algorithms for the identification of keemun black tea rank quality. Experimental results showed that all cognitive models utilizing the SFLA approach achieved steady predictive results based on eight latent variables and selected thirteen characteristic wavelength variables. The CARS-LSSVM model with the best predictive performance was proposed based on selecting ten characteristic latent variables, and the best performance indicators of the model were as follows: the root mean square error of prediction (RMSEP) was 0.0413, the correlation coefficients of prediction set (R p ) was 0.9884, and the correct discriminant rate (CDR) was 99.01% in the validation process. This study demonstrated that cognitive spectroscopy represented a proper strategy for the highly identification of quality rankings of keemun black tea., Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2020 Elsevier B.V. All rights reserved.)

Background: Preprocessing technologies of fresh tea leaves have a great influence on tea quality. A multifunctional preprocessing device for tea raw materials has been designed and utilized as a novel item of equipment to synchronously meet the process needs of spreading of green tea, withering of black tea and shaking of oolong tea., Results: The preprocessing parameters of fresh tea leaves for spreading of green tea, withering of black tea and shaking of oolong tea were optimized by orthogonal experiments. Sensory assessment combined with statistical tools was employed as an analytical method to evaluate the pretreatment effect of processing different sorts of tea. The range analysis and variance analysis of tea sensory evaluation combined with chemical components (total polyphenols, free amino acids and soluble sugar) showed that A 3 B 2 C 3 (70%, 25 °C, 8 h), A 1 B 3 C 1 (60%, 28 °C, 18 h) and A 1 B 1 C 3 (5 min, 20 °C, discontinuous leaf turning c) were considered to be the optimum schemes for the best pretreatment conditions of the above three major processing types of tea, respectively. The verification experiment of the proposed schemes was performed with satisfactory performance., Conclusion: This study demonstrated that a multifunctional preprocessing device for fresh tea leaves can be successfully applied to simultaneously estimate spreading of green tea, withering of black tea and shaking of oolong tea. © 2019 Society of Chemical Industry., (© 2019 Society of Chemical Industry.)

The appearance of tea leaves is the most intuitive factor in assessing their quality and is an important indicator of the price. The study utilizes machine vision and hyperspectral image technology to develop a rapid method for assessing the appearance quality of seven grades of Dianhong Congou black tea samples based on shape, color, texture, and feature fusion. Using the shape metrics of length, width, perimeter, narrowness, area, and rectangularity, 88 texture values, and nine color values as features, the effectiveness of intelligent recognition algorithms such as support vector machine, random forest, and least squares support vector machine (LSSVM) in classifying single features (shape, color, and texture) and fused data models of the categories of Dianhong tea samples was explored. The results showed that the accuracy of the feature fusion-based appearance quality evaluation models was all higher than the modeling results of the corresponding single features, and the accuracy of the LSSVM model based on the fused data was 100%. This study showed that feature data fusion reflects the external information of tea leaves more comprehensively and that intelligent classification algorithms combined with image analysis techniques were an effective strategy for the identification of regional black tea. [ABSTRACT FROM AUTHOR]

Summary: Food fraud causes significant economic losses for the industry and generates distrust between the consumers and traders. Tea is one of the most valued beverages throughout the world, being vulnerable to economically motivated cheat. The objective of the study was to develop the potential of hyperspectral imaging (HSI) allying multivariate analysis and data fusion to identify the authenticity of Keemun black tea quality categories. Data fusion that integrated of texture characteristics based on grey level co‐occurrence matrix and visible and near‐infrared spectral features via competitive adaptive reweighted sampling (CARS) was as the target data for modelling. Support vector machine (SVM) and random forest (RF) were utilised for the classification of tea samples of seven grades. The RF model using fused data gave the best performance with the correct discriminant rate of 92.70% for the prediction set. This study demonstrated that HSI coupled with RF was effective in identifying tea sample rank. [ABSTRACT FROM AUTHOR]

To address the problem of the lack of rapid discrimination technology for comprehensive evaluation of product quality and grade in the field of tea quality control, this study uses near-infrared spectroscopy, electronic eye, electronic tongue, and electronic nose technology to sense the external and internal quality sensing characteristics of tea leaves, and realizes comprehensive evaluation of four quality factors such as "color, aroma, taste and shape" of tea leaves through cross-sensor multimodal quality characteristics data fusion. For the first time, a comprehensive method for assessing the grade quality of tea based on deep learning algorithms combining cross-sensor multimodal fusion features such as color, texture, shape, optical, electronic tongue, and electronic nose is established. The results showed that the constructed deep convolutional neural network (CNN) classification model had a misclassification rate of only 0.86% in the prediction set, which was better than the modeling results based on traditional classification methods. This study reveals that the CNN algorithm can effectively evaluate the quality grade of tea samples and realize the digital evaluation of multiple evaluation factors of tea sensory quality. [Display omitted] • Description of tea quality using deep learning and multi-sensor feature fusion. • A comprehensive evaluation for tea "color, aroma, taste and shape" is proposed. • The constructed CNN model has a misclassification rate of only 0.86%. [ABSTRACT FROM AUTHOR]

BACKGROUND Grading represents an essential criterion for the quality assurance of black tea. The main objectives of the study were to develop a highly robust model for Chinese black tea of seven grades based on cognitive spectroscopy. RESULTS: Cognitive spectroscopy was proposed to combine near‐infrared spectroscopy (NIRS) with machine learning and evolutionary algorithms, selected feature information from complex spectral data and show the best results without human intervention. The NIRS measuring system was used to obtain the spectra of Chinese black tea samples of seven grades. The spectra acquired were preprocessed by standard normal variate transformation (SNV), multiplicative scatter correction (MSC) and minimum/maximum normalization (MIN/MAX), and the optimal pretreating method was implemented using principal component analysis combined with linear discriminant analysis algorithm. Three feature selection evolutionary algorithms, which were a genetic algorithm (GA), simulated annealing (SA) and particle swarm optimization (PSO), were compared to search the best preprocessed characteristic wavelengths. Cognitive models of Chinese black tea ranks were constructed using extreme learning machine (ELM), K‐nearest neighbor (KNN) and support vector machine (SVM) methods based on the selected characteristic variables. Experimental results revealed that the PSO–SVM model showed the best predictive performance with the correlation coefficients of prediction set (Rp) of 0.9838, the root mean square error of prediction (RMSEP) of 0.0246, and the correct discriminant rate (CDR) of 98.70%. The extracted feature wavelengths were only occupying 0.18% of the origin. CONCLUSION: The overall results demonstrated that cognitive spectroscopy could be utilized as a rapid strategy to identify Chinese black tea grades. © 2020 Society of Chemical Industry [ABSTRACT FROM AUTHOR]

The evaluation of Chinese dianhong black tea (CDBT) grades was an important indicator to ensure its quality. A handheld spectroscopy system combined with chemometrics was utilized to assess CDBT from eight grades. Both variables selection methods, namely genetic algorithm (GA) and successive projections algorithm (SPA), were employed to acquire the feature variables of each sample spectrum. A partial least‐squares discriminant analysis (PLS‐DA) and support vector machine (SVM) algorithms were applied for the establishment of the grading discrimination models based on near‐infrared spectroscopy (NIRS). Comparisons of the portable and benchtop NIRS systems were implemented to obtain the optimal discriminant models. Experimental results showed that GA‐SVM models by the handheld sensors yielded the best predictive performance with the correct discriminant rate (CDR) of 98.75% and 100% in the training set and prediction set, respectively. This study demonstrated that the handheld system combined with a suitable chemometric and feature information selection method could successfully be used for the rapid and efficient discrimination of CDBT rankings. It was promising to establish a specific economical portable NIRS sensor for in situ quality assurance of CDBT grades. [ABSTRACT FROM AUTHOR]

[Display omitted] • Digital diagnosis of Dianhong black tea quality using pocket-sized NIR. • Optimizing features by the PSO, ACO, SA, and GWO. • Recognition models for black tea categories were developed using ELM, PLS-DA, and SVM combined with the selected features. • The GWO-SVM model achieves the expected predictive performance. This paper investigates the feasibility of combining portable near-infrared (NIR) spectroscopy with chemometric algorithms to improve the quality prediction performance of Dianhong black tea. This study presented a pocket-sized NIR device (defined as NIR-S-R2) for the simultaneous identification of multiple categories of Dianhong black tea. Firstly, the spectral information of 700 multi-category Dianhong tea samples was sensed by the NIR-S-R2 system and high-quality analytical spectra of the samples were obtained by comparing multiple spectral pre-processing algorithms. Then, based on the best pre-processed spectral data, four characteristic wavelength evolution algorithms, namely particle swarm optimization, ant colony optimization, simulated annealing, and grey wolf optimization (GWO), were employed to obtain the characteristic spectral variables. Finally, extreme learning machine, partial least squares discriminant analysis, and support vector machine (SVM) methods were adopted to develop recognition models for seven categories of Dianhong black tea based on the selected features. The results showed that the GWO-SVM model had the best predictive capability, with a prediction accuracy of 91.42%. At suitable spectral wavelengths, the NIR-S-R2 system combined with chemometrics offers good discriminatory accuracy and potential for the application of extended detection. [ABSTRACT FROM AUTHOR]

Near-infrared spectroscopy and back propagation-artificial neural network (BP-ANN) model in conjunction with synergy interval partial least squares (siPLS) algorithm were used to evaluate tea shoots quality. The near-infrared spectra regions relevant to tea quality (12493 cm−1to 11645 cm−1, 9087.5 cm−1to 8242.7 cm−1, 8238.9 cm−1to 7394.2 cm−1, and 6541.7 cm−1to 5697 cm−1) were selected using siPLS algorithm. The two principal components that explained 99.46% of the variability in this spectral data were then used to calibrate the BP-ANN quality index (QI) model. The performance of this model [the coefficient of determination for prediction (), 0.9680; root mean square error of prediction (RMSEP), 0.0178] was superior to those of the BP-ANN model (= 0.9332, RMSEP = 0.0285) and the siPLS model (= 0.9230, RMSEP = 0.0360). The predicted QI values of 25 samples highly correlated with the experimental values (= 0.9223, RMSEP = 0.0344). The QI model with the combined siPLS-BP-ANN algorithms accurately predicted the quality of tea shoots. [ABSTRACT FROM PUBLISHER]

• Assessing black tea quality by Vis-NIR spectra. • Selecting spectral features by the Si-PLS, CARS, iVISSA, and VISSA. • Chemometric models for black tea quality developed using SVM-based kernel methods. • The CARS-linear kernel SVM model shows promising results. Current evaluation systems for tea quality assessment are difficult to use and unstable. A rapid and cheap assessment method is required to distinguish the quality levels of tea. In this study, a visible-near-infrared (Vis–NIR) spectrometer and support vector machine (SVM)-based kernels were used for the qualitative categorization of black tea. First, a Vis–NIR system was used to acquire the spectral data of seven levels of tea samples. The obtained spectra were preprocessed using the Savitzky–Golay smoothing combined with the first derivative and standard normal variate transformation. Then, four characteristic wavelength selection algorithms, namely synergy interval partial least square, competitive adaptive reweighted sampling (CARS), the variable iterative space shrinkage approach (VISSA), and the interval VISSA, were used to obtain preprocessed spectral features. Finally, the SVM was used with four kernel functions, namely the linear, Gaussian, quadratic, and cubic functions, to develop models based on the variables obtained from the selected features for tea quality classification. The results revealed that the CARS–linear kernel SVM model exhibited the best results, with a correct identification rate of 91.85 % in the validation process. Our findings demonstrate that Vis-NIR spectroscopy can be a rapid, inexpensive, efficient, and alternative method for predicting the quality of black tea. [ABSTRACT FROM AUTHOR]

• Assessing tea rank by Vis-NIR spectra and a hybrid variable selection strategy. • Feature data are selected by the GA, IRIV, VCPA, VCPA-GA, and VCPA-IRIV. • Authentication models for tea ranks are created using LSSVM, RF, and BPNN. • The VCPA-IRIV-LSSVM model obtains superior performance. Tea is one of the most consumed and popular beverages around the world. Authenticating tea quality is essential to resist counterfeit goods and prevent fraud. This study innovatively aims to use a visible-near infrared (Vis-NIR) sensor combined with a variable combination population analysis (VCPA)-based hybrid variable selection algorithm to realize the authentication of tea quality. First, the spectra of 700 tea samples from seven different grades are acquired using a Vis-NIR sensor. Second, the raw spectra acquired are preprocessed by Savitzky-Golay convolution smoothing combined with first derivative (SG1), standard normal variate transformation (SNV), and multiplicative scatter correction (MSC), and the best pretreating technique is determined by support vector machine (SVM) algorithm. Third, the VCPA-based hybrid method optimizes the variable space of the optimal pretreated spectra data in the first step, and a genetic algorithm (GA) and iteratively retaining informative variables (IRIV) are utilized to execute further optimization and shrink the variable in the second step. Finally, authentication models of tea quality are constructed employing least squares support vector machine (LSSVM), random forest (RF), and back propagation neural network (BPNN) approaches based on the selected variables. The results show that the VCPA-IRIV-LSSVM model obtains the best predictive performance with an accuracy of 97.40% in the predictive process. Our findings demonstrate that it is feasible to apply the inexpensive vis-NIR sensor to assess quality attributes in tea. Also, the VCPA-based hybrid method can search the combination of the optimal wavelength and improve the identification accuracy and robustness of the model and can assist sensor manufacturers in developing a cheap and user-friendly smartphone-coupled micro spectroscopic sensor for authentication of tea quality. [ABSTRACT FROM AUTHOR]

• A productive equipment for sorting famous tea raw materials was designed. • The sorting scheme consists of classification, control and process scheme. • Higher grades of fresh tea leaves have a longer horizontal throwing trajectory. • The sorting effect of high-grade raw material is better under the optimal parameters. • Improving the classification performance of the model using genetic algorithms. Famous tea is a pillar of China's tea industry and must be processed using fresh leaves of consistent tenderness and size. However, equipment with high accuracy and efficiency are unavailable for sorting fresh tea leaves (FTLs). Therefore, in this study, we designed a machine-vision platform for sorting FTLs. This platform subjects FTLs to a horizontal tossing motion and change trajectory by blowing air, thereby achieving the sorting of different types of FTLs. The overall sorting scheme consists of classification scheme, control scheme, and process scheme. We found that different types of FTLs have different flight trajectories when subjected to horizontal tossing. And have different optimal classification features and sorting parameters. When sorting high-grade FTLs (low-grade FTLs), the classification feature used is area and mean saturation (equivalent diameter, major-axis length and G variance), the designed platform achieved a recognition rate, purity, selection rate, and integrity rate of 100% (98.5%), 94.32% (87.47%), 91.67% (90.67%), and 100% (100%), respectively. Under optimal parameters, the production efficiency of the platform reached 25 kg/h. Overall, the results of this study indicate that the designed platform can meet the small-scale production requirements for famous tea. [ABSTRACT FROM AUTHOR]