Search

This study aimed to compare the moisture content (MC) uniformity of ginger slices subjected to microwave-vacuum drying (MVD) and hot-air drying (HAD) and discriminate MVD ginger slices from HAD ones (MC < 20 g/100 g) by NIR hyperspectral imaging (HSI). MC maps were visualized by a simplified partial least squares regression (PLSR) model with the determination coefficient in the prediction of 0.971. The hole effect of the semivariogram in MC maps indicated the moisture difference between the center and the edge. Furthermore, principal component analysis (PCA) of the semivariogram indicated a similar degree of moisture dissimilarity between MVD and HAD ginger slices with MC lower than 60 g/100 g. Partial least squares-discriminant analysis (PLS-DA) and support vector machines (SVM) classifiers were respectively applied to the datasets of spectral reflectance at full wavelengths and nine optimum wavelengths selected by PCA, and variances in E-W and N-S directions to build classification models. The highest correct classification rate (CCR) of 95.1% was achieved by applying PLS-DA or SVM to the dataset of full wavelengths, while CCR of 90.3% was obtained for SVM model based on the dataset of nine selected wavelengths, confirming the potential of HSI spectra for MVD and HAD ginger slices discrimination. [ABSTRACT FROM AUTHOR]

Hyperspectral imaging technology was investigated to examine the changes of sarcoplasmatic and myofibrillar proteins contents in boiled pork. Different spectral pre-processing methods were employed to improve the partial least squares regression (PLSR) models. Seven optimal wavelengths in the region of 400–2000 nm without spectral pre-processing were used to build the best PLSR model for predicting the sarcoplasmatic proteins contents with R2 p of 0.9185 and RMSEP of 2.50 mg/g. Based on multiplicative scatter correction (MSC) method, eight wavelengths in the region of 400–1000 nm were also selected to construct the best predictive PLSR model for myofibrillar protein contents with R2 p of 0.9161 and RMSEP of 2.71 mg/g. The changes in sarcoplasmatic and myofibrillar proteins contents of boiled pork during different cooking periods were successfully visualized. • Sarcoplasmatic proteins content of pork was stabilized after boiling 9 min. • Myofibrillar proteins content of pork was stabilized after boiling 20 min. • Original reflectance spectra were suitable to predict sarcoplasmatic proteins. • MSC absorbance spectra were suitable to predict sarcoplasmatic proteins. • PLSR model of sarcoplasmatic and myofibrillar proteins having accuracy above 0.91. [ABSTRACT FROM AUTHOR]

• Pathogenesis of pear infected by A. alternata was monitored by HSI. • HSI was successfully used for early identification of pear black spot disease. • Spectral angle mapper was used to identify and segment the fruit infected area. • SVM model had better classification accuracy for early detection of disease. Pathogenetic process monitoring and early detection of black spot disease caused by Alternaria alternata on pear fruit is still difficult, as it causes only minor changes in the appearance of the infected area during the early stages of infection. In this study, the potential of hyperspectral imaging (HSI) for monitoring the pathogenetic process and early detection of the disease on pear fruit was evaluated. Fresh Korla pears were inoculated with Alternaria alternata and hyperspectral images were acquired from infected and control samples. Spectral angle mapping was performed to segment the infected area from sound tissue, and to monitor the pathogenetic process of the disease. Support vector machine (SVM), k-nearest neighbor, and partial least squares discriminant analysis models were developed and evaluated for their ability to detect early onset of the disease. Results concluded that the SVM model with an overall accuracy of 97.5% was most suitable for the proposed HSI technique. This study is the first reported attempt to use HSI to monitor the pathogenetic process and detect the early symptom of the disease in pear fruit. [ABSTRACT FROM AUTHOR]

This study aimed to investigate the impact of microwave-vacuum drying (MVD) on moisture uniformity in mushroom slices and whole mushrooms (Agaricus bisporus) by near infrared (NIR) hyperspectral imaging in combination with multivariate analysis. The partial least squares regression (PLSR) models based on standard normal variate (SNV) preprocessing spectra obtained good results for predicting moisture contents (MC) with the coefficient of determination in prediction ( R p 2 ) of 0.985 for mushroom slices and 0.963 for whole mushrooms. The MC of caps was 5–11 g/100 g higher than that of stipes for the first 12 min of drying and after 18 min of drying the MC of stipes was 6–22 g/100 g higher than that of caps. The MC of peripheral caps of whole mushrooms was 2–12 g/100 g higher than that of centers for the first 30 min of drying, but the MC of peripheral caps was 4–5 g/100 g lower than that of centers in the last 20 min of drying. In addition, the difference of MC between caps and stipes (6–22 g/100 g) in mushroom slices was more significant than that between the center and periphery (4–5 g/100 g) in whole mushrooms during the middle and later periods of drying procedure. • Rates of shrinkages, color changes and evaporation vary during drying. • Caps of mushroom slices tend to lose moisture faster than stipes. • Stipes under caps of whole mushrooms can cause uneven moisture distribution. • Moisture of sliced and whole mushrooms shows different patterns during drying. [ABSTRACT FROM AUTHOR]

Near infrared (NIR) hyperspectral imaging and Fourier transform mid infrared (FT-MIR) microspectroscopy were explored in the current study to investigate how constituent elements of sweet potato change during cooking, and in the meantime, to identify sweet potato varieties. Partial least square discriminant analysis (PLSDA) model was established to classify varieties of sweet potato, and the correct classification rate of the PLSDA model using Spectral Set I (964–1645 nm) reached as high as 100%. Competitive adaptive reweighted sampling (CARS) was introduced to choose incipient feature wavelengths from three spectral subsets related to tuber cooking loss (CL). Based on 8 feature variables from Spectral Set I, CARS-SVMR model performed best with the highest coefficient of determination in prediction (R 2 P) of 0.893 and the lowest root mean square error of prediction (RMSEP) of 0.075. Then, these three subsets of feature wavelengths selected by CARS were re-optimised by using successive projections algorithm (SPA). With 7 feature variables from Spectral Set II (3996–600 cm−1) suggested by CARS-SPA, the CARS-SPA-PLSR model predicted tuber CL with R 2 P of 0.773 and RMSEP of 0.079. Moreover, the CARS-SPA-PLSR model using 5 wavelengths from Spectral Set I exhibited good prediction result, with R 2 P of 0.913 and RMSEP of 0.058. Although both techniques are capable of determining sweet potato CL in an effective way, the NIR technology demonstrates better predictive capability based on the reduced CARS-SPA-PLSR model. Highlights • IR imaging spectroscopy was proposed to characterise sweet potato samples. • Sweet potato varieties were identified by the PLSDA with NIR and MIR spectra. • PLSR and SVMR were compared using the feature variables selected by CARS-SPA. • CARS-SPA-PLSR model presented acceptable accuracy in the prediction of tuber CL. • NIR hyperspectral imaging showed better performance than MIR microspectroscopy. [ABSTRACT FROM AUTHOR]

[Display omitted] • SVC, RF and CNN models were built to distinguish beef of different states. • RFE-RF and FI-RF were used to select the optimal characteristic wavelengths. • Different fusing spectra and texture methods based on CNN were proposed and compared. • Visualisation of beef in different states was achieved. In this study, hyperspectral imaging (HSI) technology combined with a convolutional neural network (CNN) was used to distinguish fresh and frozen-thawed beef samples. After obtaining hyperspectral data of beef with different freezing/thawing cycles, the CNN was used to extract spectral features of all bands to compare with recursive feature elimination based on random forest and feature importance based on random forest. Then, eight characteristic wavelengths extracted by the first derivative-feature importance based on the random forest were used to establish the CNN model with an accuracy of 86.11%. Textural features of beef were used in the CNN model with early feature fusion of spectra and texture and late feature fusion of spectra and texture, and the CNN model using early feature fusion of spectra and texture showed more excellent results with an accuracy of 88.89%. Finally, beef samples in different states were well visualised. The research in the current study should provides a potential detection method for non-destructive and rapid tracing beef of different states. [ABSTRACT FROM AUTHOR]

This study was carried out to explore the potential of computer vision system (CVS) and two hyperspectral imaging (HSI) systems covering the visible and short-wave near infrared range (400–1000 nm) and the long-wave near infrared range (897–1753 nm), respectively, for differentiation of organic and conventional farm-raised salmon fillets in fresh and chill-stored conditions. Partial least squares-discriminant analysis (PLS-DA), support vector machine (SVM) and random forests (RF) classifiers were used to build classification models for recognition and authentication of the tested samples. The results suggested that hyperspectral discrimination performed much better than CVS. For the same validation set, the highest correct classification rate (CCR) equivalent to 98.2% was presented in two SVM models, one was built from full spectral variables in the 400–1000 nm and the other from four optimal wavelengths in the same spectral region. The best prediction for CVS was obtained by using PLS-DA with CCR of 83.6% for validation, while the most satisfying results for HSI in 897–1753 nm were achieved by applying SVM algorithm on full spectral region as well as 10 important wavelengths with both CCRs of 92.7% for validation. In a nutshell, hyperspectral imaging in the 400–1000 nm has presented the best predictive ability to differentiate between organic and conventional salmon fillets, while SVM classifier has been confirmed to be very powerful for multivariable analysis in our case. Overall, the outcome achieved from this research strongly suggested the capability of the hyperspectral imaging for objective and rapid categorization of the two salmon varieties under circumstances of fresh and chill-stored conditions. [ABSTRACT FROM AUTHOR]

This study aimed to develop an alternative technique for rapid, accurate and non-invasive determination of gross energy density values of salmon fillets based on hyperspectral imaging (900–1700 nm). Spectral data were extracted from the hyperspectral images with outlier pixels removed. Good performances were achieved with r P of 0.906 and 0.909, and normalized RMSEP of 6.768% and 7.064% for partial least squares regression (PLSR) and epsilon-support vector regression (epsilon-SVR) analysis, respectively. The optimized stepwise-PLS model built with four wavelengths (931, 1001, 1135 and 1168 nm) yielded good results with r P of 0.908 and normalized RMSEP of 6.871%. The distribution maps for visualizing energy density indices in all portions of the salmon fillet were subsequently generated. The overall results confirmed the successful development of a rapid and non-destructive technique using hyperspectral imaging for determining energy value of salmon fillets, and the provision of an alternative method for caloric density measurement for the food industry. [ABSTRACT FROM AUTHOR]

Chlorophyll a (Chl-a) is regarded as one of the important components to estimate water quality and sustainability of freshwater aquaculture operations. In the current study, a hyperspectral imaging (HSI) system was used to determine the effect of season models on the accuracy of Chl-a estimation in outdoor aquaculture ponds. A visible and near infrared hyperspectral imaging system (400–1000 nm) was used to measure surface spectral reflectance ( R ) of water collected from outdoor ponds in four different seasons. Firstly, values of surface spectral reflectance ( R ) were amplified by a baseline correction (740 nm). Two-band, three-band and four-band spectral reflectance were used to compute Chl-a concentration and a new cross band ratio algorithm with six wavelengths was proposed in the study. Results indicated that two-band model established based on reflectance ratio ( R 702 /R 666 ) had better performances for Chl-a prediction with determination coefficients ( r 2 ) of 0.908 than those by ( R 675 - 1 − R 691 - 1 )* R 743 and ( R 675 - 1 − R 691 - 1 )/( R 743 - 1 − R 691 - 1 ) models with r 2 of 0.902 and 0.896, respectively. Six optimal wavelengths (410, 682, 691, 966, 972, and 997) were identified using successive projections algorithm (SPA). The optimized regression model ( R 410 - 1 − R 966 - 1 )/( R 682 - 1 − R 972 - 1 )/( R 691 - 1 − R 997 - 1 ) showed best result with r 2 of 0.961 for Chl-a prediction. Model of cross band ratio algorithm with six wavelengths was mapped to each pixel in the image to display Chl-a component in outdoor ponds under four different seasons. The current study showed that it was feasible to use the HSI system for monitoring the influence of seasons for outdoor aquaculture water quality. [ABSTRACT FROM AUTHOR]

This study investigated the potential of using hyperspectral imaging technology (900–1700 nm) to determine the thiobarbituric acid (TBA) value and pH for evaluation of lipid oxidation in Atlantic salmon ( Salmo salar ) fillets during cold storage for 0, 3, 6, 9 and 12 days at 1 ± 1 °C. Good results were achieved for both parameters by using partial least square regression (PLSR) calibration models with full spectral region. Two simplified models were then built by using forward stepwise-multiple linear regression (MLR) variable selection method to select 18 and 10 most important wavelengths for TBA value and pH, respectively. The optimised stepwise-MLR model for TBA value yielded satisfactory results with correlation coefficient ( r C ) of 0.921 and root mean square error of calibration (RMSEC) of 1.840 μmol MDA/kg fish. This model was used to visualise the TBA value distributions during different storage days. Further improvements were achieved by applying particle analysis on the images to extract only the spectra from white stripes in salmon fillet. When in tandem with detrend pre-processing technique, the calibration model based on particle analysis demonstrated the best performance for TBA value prediction ( r C = 0.957 and RMSEC = 1.449 μmol MDA/kg). In addition, a novel chemometric strategy accomplished by the use of hyperspectrograms was also proposed in this work and satisfactory results were obtained. The overall results confirmed the capability of near-infrared hyperspectral imaging as a rapid and non-invasive technique to monitor lipid oxidation in salmon fillets. [ABSTRACT FROM AUTHOR]

Food quality and safety is the foremost issue for consumers, retailers as well as regulatory authorities. Most quality parameters are assessed by traditional methods, which are time consuming, laborious, and associated with inconsistency and variability. Non-destructive methods have been developed to objectively measure quality attributes for various kinds of food. In recent years, hyperspectral imaging (HSI) has matured into one of the most powerful tools for quality evaluation of agricultural and food products. HSI allows characterization of a sample's chemical composition (spectroscopic component) and external features (imaging component) in each point of the image with full spectral information. In order to track the latest research developments of this technology, this paper gives a detailed overview of the theory and fundamentals behind this technology and discusses its applications in the field of quality evaluation of agricultural products. Additionally, future potentials of HSI are also reported. [ABSTRACT FROM AUTHOR]

Currently, the issue of food safety and quality is a great public concern. In order to satisfy the demands of consumers and obtain superior food qualities, non-destructive and fast methods are required for quality evaluation. As one of these methods, hyperspectral imaging (HSI) technique has emerged as a smart and promising analytical tool for quality evaluation purposes and has attracted much interest in non-destructive analysis of different food products. With the main advantage of combining both spectroscopy technique and imaging technique, HSI technique shows a convinced attitude to detect and evaluate chicken meat quality objectively. Moreover, developing a quality evaluation system based on HSI technology would bring economic benefits to the chicken meat industry. Therefore, in recent years, many studies have been conducted on using HSI technology for the safety and quality detection and evaluation of chicken meat. The aim of this review is thus to give a detailed overview about HSI and focus on the recently developed methods exerted in HSI technology developed for microbiological spoilage detection and quality classification of chicken meat. Moreover, the usefulness of HSI technique for detecting fecal contamination and bone fragments of chicken carcasses are presented. Finally, some viewpoints on its future research and applicability in the modern poultry industry are proposed. [ABSTRACT FROM PUBLISHER]

Nowadays, people have increasingly realized the importance of acquiring high quality and nutritional values of fish and fish products in their daily diet. Quality evaluation and assessment are always expected and conducted by using rapid and nondestructive methods in order to satisfy both producers and consumers. During the past two decades, spectroscopic and imaging techniques have been developed to nondestructively estimate and measure quality attributes of fish and fish products. Among these noninvasive methods, visible/near-infrared (VIS/NIR) spectroscopy, computer/machine vision, and hyperspectral imaging have been regarded as powerful and effective analytical tools for fish quality analysis and control. VIS/NIR spectroscopy has been widely applied to determine intrinsic quality characteristics of fish samples, such as moisture, protein, fat, and salt. Computer/machine vision on the other hand mainly focuses on the estimation of external features like color, weight, size, and surface defects. Recently, by incorporating both spectroscopy and imaging techniques in one system, hyperspectral imaging cannot only measure the contents of different quality attributes simultaneously, but also obtain the spatial distribution of such attributes when the quality of fish samples are evaluated and measured. This paper systematically reviews the research advances of these three nondestructive optical techniques in the application of fish quality evaluation and determination and discuss future trends in the developments of nondestructive technologies for further quality characterization in fish and fish products. [ABSTRACT FROM PUBLISHER]

The potential of visible and near infrared (400–1000 nm) hyperspectral imaging as a rapid and non-invasive method was investigated to differentiate freshness of prawns. In both unfrozen and frozen groups (a total of 280 prawns), two different freshness levels were used for classification, respectively. Mean spectral data from the full surface of prawns were extracted automatically as the hyperspectral cubes. Both the first and second derivative spectra were performed for waveform analysis. Successive projections algorithm (SPA) was conducted to select the individual feature wavelengths for classification. Least squares-support vector machine (LS-SVM), adaptive boosting (AdaBoost) algorithm and back-propagation neutral networks (BP-NN) were carried out for classification using the derivative spectrums based on both full wavelengths and selected feature wavelengths. The results demonstrated that SPA–LS-SVM achieved satisfactory average correct classification rate of 98.33% and 95% for prediction samples in unfrozen and frozen groups, respectively. Visualization map of classification of eight prawns (two groups) was also presented. The overall results revealed that hyperspectral imaging technique is promising for freshness classification of prawns rapidly and non-invasively. [ABSTRACT FROM AUTHOR]

Fresh and frozen-thawed (F-T) pork meats were classified by Vis–NIR hyperspectral imaging. Eight optimal wavelengths (624, 673, 460, 588, 583, 448, 552 and 609 nm) were selected by successive projections algorithm (SPA). The first three principal components (PCs) obtained by principal component analysis (PCA) accounted for over 99.98% of variance. Gray-level-gradient co-occurrence matrix (GLGCM) was applied to extract 45 textural features from the PC images. The correct classification rate (CCR) was employed to evaluate the performance of the partial least squares-discriminate analysis (PLS-DA) models, by using (A) the reflected spectra at full wavelengths and (B) those at the optimal wavelengths, (C) the extracted textures based on the PC images, and (D) the fused variables combining spectra at the optimal wavelengths and textures. The results showed that the best CCR of 97.73% was achieved by applying (D), confirming the high potential of textures for fresh and F-T meat discrimination. [ABSTRACT FROM AUTHOR]

Application of wavelet analysis to near-infrared (NIR) hyperspectral imaging data was exploited for categorization of lamb muscles in this study. A variety of common wavelet transforms was investigated to identify the best wavelet features for categorization of lamb muscles. The fifth-order Daubechies wavelet ('db5') was found to be the best wavelet function for decomposition of lamb spectral signal. Features of wavelet coefficients extracted from db5 wavelet at the fifth decomposition level were then used as the inputs of least-squares support vector machine (LS-SVM) for developing classification models. Principal component analysis (PCA) was used for dimensionality reduction. Classification performance of LS-SVM classifiers in tandem with wavelet transform and PCA was compared with the LS-SVM models based on original, first derivative, second derivative, smoothing, standard normal variate (SNV), and multiplicative scatter correction (MSC) spectral data; then, the overall correct classification performance for the training and test sets using combination with wavelet approximation and detail coefficients in fifth decomposition scale and PCA was 100 and 96.15 %, respectively. In addition, the developed classification models were successfully applied to the hyperspectral images for obtaining classification maps and the kappa coefficient of 0.83 was obtained for the visual classification. The results revealed that the application of wavelet analysis has a great potential for categorization of lamb muscles in tandem with multivariate analysis and image processing. [ABSTRACT FROM AUTHOR]

Highlights: [•] Spectra and texture was combined to predict pH in salted meat. [•] GLGCM was used for texture extraction from PC1 score image. [•] Feature level fusion and mean normalization was used for data fusion. [•] Model based on data fusion was superior to that based on spectra or texture alone. [•] R 2 P of 0.794 for the pH prediction based on data fusion was achieved. [ABSTRACT FROM AUTHOR]

Highlights: [•] Hyperspectral imaging technique was firstly used to predict the texture of prawns. [•] Two prediction models based on PLSR and LS-SVM were established. [•] Successive projections algorithm was conducted to select the optimal wavelengths. [•] A non-linear relationship was found between the spectrum and texture property. [•] The distribution maps of textural features were created and visualized. [Copyright &y& Elsevier]

Highlights: [•] Hyperspectral imaging was used to determine drip loss and pH in salmon fillets. [•] The predictive abilities of PLSR models with two spectral ranges were compared. [•] Important wavelengths were selected to reduce redundancy of hyperspectral images. [•] Multivariate models were built for predicting drip loss and pH with good accuracy. [•] We developed image processing algorithm to obtain drip loss and pH distribution maps. [ABSTRACT FROM AUTHOR]

Highlights: [•] Hyperspectral imaging was used to assess and predict tenderness in salmon fillets. [•] PLSR and LS-SVM models were developed to predict the WBSF in samples. [•] Optimal wavelengths were selected by means of RC and SPA methods. [•] Good accuracy of multivariate models was obtained for predicting WBSF. [•] An image processing algorithm was developed to create WBSF distribution map. [ABSTRACT FROM AUTHOR]

Abstract: Textural firmness is a primary determinant of consumer acceptance for evaluating freshness quality of fish fillet flesh. The objective of this study was to investigate the potential of using visible and near-infrared hyperspectral imaging (400–1000nm) for non-destructive prediction of firmness quality of grass carp fillet as affected by frozen storage. Fillet samples were frozen at −20°C for 24h and then stored at 4°C for thawing over five days. Hyperspectral images were obtained at different thawing stages and their corresponding spectral data were extracted. Two calibration models were established between the extracted spectral data and the reference firmness values measured by the traditional mechanical method by using partial least squares regression (PLSR) and least-square support vector machine (LS-SVM) analysis. Three approaches of regression coefficients (RC) from PLSR analysis, genetic algorithm (GA) and successive projection algorithm (SPA) were utilized to recognize the most important wavelengths that possessed the greatest influence and sensitivity on the firmness prediction based upon the whole spectral range. By comparing the above-mentioned three variable selection methods, seven optimal wavelengths (450, 530, 550, 616, 720, 955 and 980nm) were selected by GA and its corresponding simplified prediction model of GA-LS-SVM was also obtained, showing the best performance with the highest determination coefficient (R2 P) of 0.941 and the lowest root mean square error estimated by prediction (RMSEP) of 1.229. The overall results of this study suggested that hyperspectral imaging technique has a potential for fast and non-destructive prediction and analysis of textural firmness of grass carp fillets as affected by frozen storage. [Copyright &y& Elsevier]

Abstract: Hyperspectral imaging (HSI) has been previously used as a non-destructive technology to accurately determine water content in porcine meat at the initial stages of the salting process. However, since water content alone is not sufficient to predict the shelf-life and quality of meat products. In this study, prediction models for salt content and water activity (aw) of meat slices during salting process were developed. Hyperspectral images (400–1000nm) of pork slices were acquired at different periods of salting process. Spectral information was used for predicting salt content and aw by partial least square regression (PLSR). The models yielded acceptable results with determination coefficients in prediction (R2 p) of 0.928 and 0.909 for NaCl content and aw, respectively. On the basis of the optimal wavelengths identified using weighted regression coefficients from the PLSR models, three linear calibration algorithms including PLSR, principal component regression (PCR) and multiple linear regression (MLR) were compared. The optimized regression models had good performances with MLR model achieving R2 p of 0.930 and 0.914 as well as root mean squared errors for prediction (RMSEP) of 0.682 and 0.007 for prediction of salt content and aw, respectively. Distributional analysis of salt content was realized by transferring the quantitative model to each pixel in the image to display salt uptake in pork slices at different salting periods. The prediction models were accurate enough to consider HSI as a useful tool for controlling and optimizing the meat salting process. Industrial relevance: This feasibility study demonstrated that hyperspectral imaging offers the possibility for process monitoring and control and for optimization of key parameters during salting process in the meat industry. [Copyright &y& Elsevier]

Abstract: Bacterial pathogens are the main culprits for outbreaks of food-borne illnesses. This study aimed to use the hyperspectral imaging technique as a non-destructive tool for quantitative and direct determination of Enterobacteriaceae loads on chicken fillets. Partial least squares regression (PLSR) models were established and the best model using full wavelengths was obtained in the spectral range 930–1450nm with coefficients of determination R 2 ⩾0.82 and root mean squared errors (RMSEs) ⩽0.47log10 CFUg−1. In further development of simplified models, second derivative spectra and weighted PLS regression coefficients (BW) were utilised to select important wavelengths. However, the three wavelengths (930, 1121 and 1345nm) selected from BW were competent and more preferred for predicting Enterobacteriaceae loads with R 2 of 0.89, 0.86 and 0.87 and RMSEs of 0.33, 0.40 and 0.45log10 CFUg−1 for calibration, cross-validation and prediction, respectively. Besides, the constructed prediction map provided the distribution of Enterobacteriaceae bacteria on chicken fillets, which cannot be achieved by conventional methods. It was demonstrated that hyperspectral imaging is a potential tool for determining food sanitation and detecting bacterial pathogens on food matrix without using complicated laboratory regimes. [Copyright &y& Elsevier]

Abstract: This study was conducted to explore the potential of hyperspectral imaging (HSI) technique in visible and near-infrared (VIS–NIR) region (400–1700nm) for determining the spatial distribution of moisture content (MC) in farmed Atlantic salmon fillets rapidly and non-destructively. The quantitative relationship between spectral data and the reference MC values was successfully established by partial least squares regression (PLSR). Three spectral ranges of 400–1000nm (Spectral Range I), 900–1700nm (Spectral Range II) and 400–1700nm (Spectral Range III) were considered, and their results were compared to choose the best spectral range. The established PLSR models had coefficients of determination (R P 2) of 0.893, 0.902 and 0.849, and root-mean-square errors of prediction (RMSEP) of 1.513%, 1.450% and 1.800% for three spectral ranges, respectively. Important wavelengths were then selected by using regression coefficients of PLSR models for three spectral ranges, and optimised PLSR models were built using only the important wavelengths, resulting in R P 2 of 0.893, 0.888 and 0.884 with RMSEP of 1.517%, 1.553% and 1.578% for three spectral ranges, respectively. PLSR model with eight important wavelengths (420, 445, 545, 585, 635, 870, 925 and 955nm) selected from Spectral Range I was considered as the best model for MC determination and was transferred to each pixel within the image for visualising MC in all locations of salmon fillets with an aid of a developed image processing algorithm. The results revealed that hyperspectral imaging technique has a great potential to predict the MC distribution of salmon fillets non-destructively and accurately. In addition to realising the MC difference within salmon fillets, it could be possible for hyperspectral imaging to classify and grade salmon fillets based on different MC levels. Industrial Relevance: The results revealed that hyperspectral imaging technique has a great potential to predict the MC distribution of salmon fillets non-destructively and accurately for the food industry. [Copyright &y& Elsevier]

Abstract: Currently, the time of grape harvest is normally determined according to the sugar level in the pulp of the berry. Nonetheless, the stage of maturation in grape seeds should be taken into account more frequently to decide the appropriate harvest period. There are chemical and sensory analyses available to assess stage of maturation of grape seeds but they are destructive and time-consuming. Hyperspectral imaging is an alternative technology to characterize the grape seeds according to their chemical attributes, and the current work aimed to non-destructively characterize grape seeds in regard of the variety and stage of maturation. For this purpose, 56 samples of seeds from two red grape varieties (Tempranillo and Syrah) and one white variety (Zalema) in two kinds of soil were selected to assess their features based on the reflectance in the near-infrared (NIR) spectra by using prediction models (partial least squares regression) and multivariate analysis methods (principal component analysis and general discriminant analysis). In this study, a reliable methodology for predicting the stage of maturation was developed, and it was shown that it was possible to distinguish the variety of grape and even the type of soil from hyperspectral images of grape seeds. [Copyright &y& Elsevier]

Abstract: Temperature fluctuation during cold storage of meat products usually leads to undesirable microbial growths, which affect the overall product quality. In this study, a pushbroom hyperspectral imaging system in the near-infrared (NIR) range (900–1700nm) as a rapid and non-destructive technique was exploited for determining the total viable count (TVC) and psychrotrophic plate count (PPC) in chilled pork during storage. Fresh pork samples from the longissimus dorsi muscle were obtained directly from a commercial slaughtering plant, and stored in the refrigerated temperatures at 0°C and 4°C for 21days. Every 48h, a NIR hyperspectral image in the reflectance mode was acquired directly for each sample. The TVC and PPC were determined simultaneously by classical microbiological plating methods and multivariate statistical models for predicting contamination and spoilage conditions in the samples were then developed. Partial least squares regression (PLS) was applied to fit the spectral information extracted from the samples to the logarithmic values of TVC and PPC. The best regressions were obtained with R 2 of 0.86 and 0.89 for log (TVC) and log (PPC), respectively. The most important wavelengths were then selected for regression and for spatial visualization of contamination. Results are encouraging and show the promising potential of hyperspectral technology for detecting bacterial spoilage in pork and tracking the increase of microbial growth of chilled pork during storage at different temperatures. Industrial relevance: A novel method based on hyperspectral imaging technique has been successfully developed for determining the total viable count (TVC) and psychrotrophic plate count (PPC) in chilled pork during storage non-destructively for the meat industry. [Copyright &y& Elsevier]

The requirements of reliability, expeditiousness, accuracy, consistency, and simplicity for quality assessment of food products encouraged the development of non-destructive technologies to meet the demands of consumers to obtain superior food qualities. Hyperspectral imaging is one of the most promising techniques currently investigated for quality evaluation purposes in numerous sorts of applications. The main advantage of the hyperspectral imaging system is its aptitude to incorporate both spectroscopy and imaging techniques not only to make a direct assessment of different components simultaneously but also to locate the spatial distribution of such components in the tested products. Associated with multivariate analysis protocols, hyperspectral imaging shows a convinced attitude to be dominated in food authentication and analysis in future. The marvellous potential of the hyperspectral imaging technique as a non-destructive tool has driven the development of more sophisticated hyperspectral imaging systems in food applications. The aim of this review is to give detailed outlines about the theory and principles of hyperspectral imaging and to focus primarily on its applications in the field of quality evaluation of agro-food products as well as its future applicability in modern food industries and research. [ABSTRACT FROM AUTHOR]

During the last two decades, a number of methods have been developed to objectively measure meat quality attributes. Hyperspectral imaging technique as one of these methods has been regarded as a smart and promising analytical tool for analyses conducted in research and industries. Recently there has been a renewed interest in using hyperspectral imaging in quality evaluation of different food products. The main inducement for developing the hyperspectral imaging system is to integrate both spectroscopy and imaging techniques in one system to make direct identification of different components and their spatial distribution in the tested product. By combining spatial and spectral details together, hyperspectral imaging has proved to be a promising technology for objective meat quality evaluation. The literature presented in this paper clearly reveals that hyperspectral imaging approaches have a huge potential for gaining rapid information about the chemical structure and related physical properties of all types of meat. In addition to its ability for effectively quantifying and characterizing quality attributes of some important visual features of meat such as color, quality grade, marbling, maturity, and texture, it is able to measure multiple chemical constituents simultaneously without monotonous sample preparation. Although this technology has not yet been sufficiently exploited in meat process and quality assessment, its potential is promising. Developing a quality evaluation system based on hyperspectral imaging technology to assess the meat quality parameters and to ensure its authentication would bring economical benefits to the meat industry by increasing consumer confidence in the quality of the meat products. This paper provides a detailed overview of the recently developed approaches and latest research efforts exerted in hyperspectral imaging technology developed for evaluating the quality of different meat products and the possibility of its widespread deployment. [ABSTRACT FROM PUBLISHER]

Abstract: Many subjective assessment methods for fresh meat quality are still widely used in the meat industry, making the development of an objective and non-destructive technique for assessing meat quality traits a vital need. In this study, a hyperspectral imaging technique was investigated for objective determination of pork quality attributes. Hyperspectral images in the near infrared region (900–1700nm) were acquired for pork samples from the longissimus dorsi muscle, and the representative spectral information was extracted from the loin eye area. Several mathematical pre-treatments including first and second derivatives, standard normal variate (SNV) and multiplicative scatter correction (MSC) were applied to examine the influence of spectral variations in predicting pork quality characteristics. Spectral information was used for predicting color features (L, a, b, chroma and hue angle), drip loss, pH and sensory characteristics by partial least-squares regression (PLS-R) models. Independent sets of feature-related wavelengths were selected for predicting each quality attribute. The results showed that color reflectance (L), pH and drip loss of pork meat could be predicted with determination coefficients of 0.93, 0.87 and 0.83, respectively. The regression coefficients from the PLS-R models at the selected optimal wavelengths were applied in a pixel-wise manner to convert spectral images to prediction maps that display the distribution of attributes within the sample. Results indicated that this technique is a potential tool for rapid assessment of pork quality. [Copyright &y& Elsevier]

Abstract: In this study, a hyperspectral imaging technique was developed to achieve fast, accurate, and objective determination of pork quality grades. Hyperspectral images were acquired in the near-infrared (NIR) range from 900 to 1700nm for 75 pork cuts of longissimus dorsi muscle from three quality grades (PSE, RFN and DFD). Spectral information was extracted from each sample and six significant wavelengths that explain most of the variation among pork classes were identified from 2nd derivative spectra. There were obvious reflectance differences among the three quality grades mainly at wavelengths 960, 1074, 1124, 1147, 1207 and 1341nm. Principal component analysis (PCA) was carried out using these particular wavelengths and the results indicated that pork classes could be precisely discriminated with overall accuracy of 96%. Algorithm was developed to produce classification maps of the tested samples based on score images resulting from PCA and the results were compared with the ordinary classification method. Investigation of the misclassified samples was performed and showed that hyperspectral based classification can aid in class determination by showing spatial location of classes within the samples. [Copyright &y& Elsevier]

Abstract: This study was carried out to develop a hyperspectral imaging system in the near infrared (NIR) region (900–1700nm) to assess the quality of cooked turkey hams of different ingredients and processing parameters. Hyperspectral images were acquired for ham slices originated from each quality grade and then their spectral data were extracted. Spectral data were analyzed using principal component analysis (PCA) to reduce the high dimensionality of the data and for selecting some important wavelengths. Out of 241 wavelengths, only eight wavelengths (980, 1061, 1141, 1174, 1215, 1325, 1436 and 1641nm) were selected as the optimum wavelengths for the classification and characterization of turkey hams. The data analysis showed that it is possible to separate different quality turkey hams with few numbers of wavelengths on the basis of their chemical composition. The results revealed the potentiality of NIR hyperspectral imaging as an objective and non-destructive method for the authentication and classification of cooked turkey ham slices. [ABSTRACT FROM AUTHOR]

Highlights • A polarized-HSI system (400–1000 nm) was developed for agricultural application. • Superficial information contributes more in discriminating sunflower varieties. • Both surface and subsurface features are equally important in classification. • R ⊥ is most important in detecting diseases (PM and/or SLS) on sunflower leaves. • Best classification model for disease detection had CCR of 0.963 for prediction. Abstract This study aims to investigate the potential of an original polarized hyperspectral imaging (HSI) setup in the spectral domain of 400–1000 nm for sunflower leaves in real-world. Dataset 1 includes hypercubes of sunflower leaves in two varieties with different life growth stages, while Dataset 2 is comprised of healthy and contaminated sunflower leaves suffering from powdery mildew (PM) and/or septoria leaf spot (SLS). Cross polarised (R ⊥), parallel polarised (R | |) reflectance signals, R BS (R | | + R ⊥) and R SS (R | | - R ⊥) spectra were obtained and used to develop partial least squares-discriminant analysis (PLS-DA) models. Surface information played an important role in separating two varieties of leaves due to the fact that the best model performance was achieved by using R SS mean spectra, while both surface and subsurface were equally important in classifying leaves between two major growth stages because model of R BS mean spectra outperformed other models. The best classification model for disease detection was achieved by using pixel R ⊥ spectra with the correct classification rate (CCR) of 0.963 for both cross validation and prediction, meaning that subsurface spectral features were the most important to detect infected leaves. The resulting classification maps were also displayed to visualize the distribution of the infected regions on the leaf samples. The overall results obtained in this research showed that the developed polarized-HSI system coupled with multivariate analysis has considerable promise in agricultural real-world applications. [ABSTRACT FROM AUTHOR]

Microwave vacuum drying of banana slices was investigated in this study. Near-infrared hyperspectral imaging (NIR-HSI, 950-1650 nm with 7 nm increments) and chemometrics were used to predict moisture content, hardness, and fracturability of banana slices during the entire drying process. A computer vision system was also used to evaluate browning index (BI) of slices at different drying stages; dark brown spots observed in slices were subjected to over 21-min total microwave heating. In addition, thermal imaging results showed that higher temperature occurred in the central part of banana slices. Calibration models were developed based on partial least square regression and support vector machine regression (SVM-R) using mean spectra with standard normal variate pretreatment. The results showed that SVM-R models performed better in the prediction of the three quality attributes with of 0.996, 0.927, and 0.961, respectively. The moisture distribution maps were generated using NIR-HSI; uniformly distributed moisture patterns were observed in each drying batch with small variations observed within individual slices. This study demonstrated the potential of spectral and imaging techniques for nondestructive quality evaluation of food products during drying processes. [ABSTRACT FROM AUTHOR]

[Display omitted] • Hyperspectral imaging (HSI) was first time used to study vacuum cooling (VC) process. • Quality attributes of mushroom were dynamically evaluated during (VC) by HSI. • Partial least squares regression models realized good prediction for the attributes. • Pseudocolour maps were developed to visualize the distribution of the attributes. • The study provides a rapid and real-time nondestructive detection strategy. Visible near-infrared (400–1000 nm) hyperspectral imaging technology (HSI) was first time used to evaluate changes in various quality attributes including moisture contents (MC), lightness (L*), yellowness (b*), hardness (H) and chewiness (ch) of Agaricus bisporus as affected by vacuum cooling (VC). Partial least squares regression (PLSR) models based on Savitzky-Golay (SG) smoothing combined with the first derivative (SG-1st der) spectral preprocessing achieved good results for predicting MC with R2 P of 0.823 and RMSEP of 0.341%. Similarly, PLSR models based on full wavelengths achieved good results for predicting L*, b* and H, and ch with R2 P of 0.796, 0.848 and 0.810, and 0.903, with RMSEP of 1.253, 0.808 and 9.254, and 2.717. Finally, pseudocolour maps were developed to visualize the distribution of these attributes during VC. The current study provides a rapid, real-time and non-destructive detection strategy for evaluating the VC process. [ABSTRACT FROM AUTHOR]

Hyperspectral imaging is built with the aggregation of imaging, spectroscopy and radiometric techniques. This technique observes the sample behaviour when it is exposed to light and interprets the properties of the biological samples. As hyperspectral imaging helps in interpreting the sample at the molecular level, it can distinguish very minute changes in the sample composition from its scatter properties. Hyperspectral data collection depends on several parameters such as electromagnetic spectrum wavelength range, imaging mode and imaging system. Spectral data acquired using a hyperspectral imaging system contain variations due to external factors and imaging components. Moreover, food samples are complex matrices with conditions of surface and internal heterogeneities, which may lead to variations in acquired data. Hence, before extracting information, these variations and noises must be reduced from the data using reference-dependent or reference-independent spectral pre-processing techniques. Using of the entire hyperspectral data for information extraction is tedious and time-consuming. In order to overcome this, exploratory data analysis techniques are used to select crucial wavelengths from the excessive hyperspectral data. Using appropriate chemometric techniques (supervised or unsupervised learning techniques) on this pre-processed hyperspectral data, qualitative or quantitative information from sample can be obtained. Qualitative information for analysing of the chemical composition, detecting of the defects and determining the purity of the food product can be extracted using discriminant analysis techniques. Quantitative information including variation in chemical constituents and contamination levels in food and agricultural sample can be extracted using categorical regression techniques. In combination with appropriate spectra pre-processing and chemometric technique, hyperspectral imaging stands out as an advanced quality evaluation system for food and agricultural products. [ABSTRACT FROM AUTHOR]

The potential of visible and near infrared (VIS/NIR) hyperspectral imaging was investigated as a rapid and nondestructive technique to determine whether fish has been frozen-thawed. A total of 108 halibut ( Psetta maxima) fillets were studied, including 48 fresh and 60 frozen-thawed (F-T) samples. Regarding the F-T samples, two speeds of freezing (fast and slow) were tested. The hyperspectral images of fillets were captured using a pushbroom hyperspectral imaging system in the spectral region of 380 to 1,030 nm. All images were calibrated for reflectance, followed by the minimum noise fraction rotation to reduce the noise. A region-of-interest (ROI) at the image center was selected, and the average spectral data were generated from the ROI image. Dimension reduction was carried out on the ROI image by principal component analysis. The first three principal components (PCs) explained over 98 % of variances of all spectral bands. Gray-level co-occurrence matrix analysis was implemented on the three PC images to extract 36 textural feature variables in total. Least squares-support vector machine classification models were developed to differentiate between fresh and F-T fish based on (1) spectral variables; (2) textural variables; (3) combined spectral and textural variables, respectively. Satisfactory average correct classification rate of 97.22 % for the prediction samples based on (3) was achieved, which was superior to the results based on (1) or (2). The results turned worse when different freezing rates were taken into consideration to classify three groups of fish. The overall results indicate that VIS/NIR hyperspectral imaging technique is promising for the reliable differentiation between fresh and F-T fish. [ABSTRACT FROM AUTHOR]

• MVD mushroom stipes showed higher moisture contents, ε ' and ε ' ' values than caps. • MVD mushroom caps displayed higher water evaporation and rehydration rates. • The centre area of the MVD mushroom slice had the highest temperature. • The more porous structure in the MVD cap led to its higher mass transfer rates. This research work aimed to investigate the effects of microstructures, dielectric property and temperature distributions on drying feature difference between the mushroom cap and stipe during the microwave-vacuum drying (MVD) process. Near-infrared hyperspectral imaging (NIR HSI) was employed to visualize distribution maps for moisture content (MC), dielectric constant ε ' and dielectric loss factor ε ' ' of mushroom slices during the MVD process. Infrared (IR) thermal imaging was used to evaluate the temperature distribution of the mushroom slices. Results demonstrated higher MC, ε ' and ε ' ' values in MVD mushroom stipes. Nevertheless, the centre area of the mushroom slice showed the highest temperature, while the MVD mushroom cap displayed a more porous structure. The effect of microstructure could encounter effects of dielectric properties and temperature to cause higher water evaporation in the MVD cap. This work highlights the novelty to combine different detection techniques to investigate the effects of microstructures, dielectric property and temperature distributions on drying patterns of mushroom slices. [ABSTRACT FROM AUTHOR]

Highlights: [•] NIR imaging has potential to authenticate and classify Spanish cooked hams. [•] Multivariate analyses using PLSR and PLS-DA were applied to develop statistical models. [•] With a few numbers of wavelengths it is possible to identify ham qualities. [•] The PLS-DA model using 966, 1061, 1148, 1256, 1373 and 1628nm resulted successful. [Copyright &y& Elsevier]

Abstract: In this study, the reliability and accuracy of hyperspectral imaging technique in tandem with multivariate analyses were investigated for identification and authentication of different red meat species. Hyperspectral images were acquired from longissimus dorsi muscle of pork, beef and lamb and their spectral data were extracted and analyzed by principal component analysis (PCA) and partial least squares discriminant analysis (PLS-DA) for recognition and authentication of the tested meat. The spectra were pre-treated by second derivative and six wavelengths (957, 1071, 1121, 1144, 1368 and 1394nm) were identified as important wavelengths from the 2nd derivative spectra. The resulting wavelengths were used in a pattern recognition algorithms for classification of meat samples with PLS-DA yielding 98.67% overall classification accuracy in the validation sets. The developed classification algorithms were then successfully applied in the independent testing set for the authentication of minced meat. The results clearly showed that the combination of hyperspectral imaging, multivariate analysis and image processing has a great potential as an objective and rapid method for identification and authentication of red meat species. Industrial Relevance: This study was carried out to investigate the potential of NIR hyperspectral imaging system for identification and authentication of red meat species for the meat industry. [Copyright &y& Elsevier]