Search

International audience; Drip irrigation is one of the most efficient irrigation techniques, but it is susceptible to dripper clogging. This study proposes a novel and non-destructive method based on visible and near infrared (Vis/NIR) spectroscopy coupled with chemometric methods for the discrimination and thickness estimation of physical and chemical fouling in drip-irrigation systems. Four representative materials linked to physical and chemical clogging (kaolin, bentonite, sand and calcium carbonate) at different thicknesses were selected to illustrate the potential of the approach. Partial least squares regression (PLSR) and its modification partial least squares with discriminant analysis (PLS-DA) were selected for the modelling of clogging materials. The PLS-DA model was able to predict with 96.97% accuracy all classes of materials. The PLSR models were able to estimate fouling thickness with relative prediction errors comprised between 134 μm and 164 μm. This difference appears mainly to be due to the physical properties of the selected materials. This prediction accuracy enabled the estimation of the clogging thickness between 10 and 21% of dripper channel coverage depending on the dripper channel section and the material under study. The proposed method offers an appropriate approach for clogging studies in drip irrigation systems that could be transferred to field applications.

In this work, two methods, which were visible near-infrared spectroscopy (VNIRS) and visible near-infrared spectroscopy combined with colorimetric sensor array (VNIRS-CSA), were used to identify the volatile compound changes of rice samples stored for 0 to 6 months. Principal component analysis (PCA), interval partial least squares (iPLS), and synergy interval partial least squares (SiPLS) were used for qualitative classification. A prediction model was established by linear discriminant analysis (LDA), which was compared with the traditional VNIRS detection technology. The results revealed that the VNIRS-CSA got better performance than VNIRS and exhibited a good result based on iPLS/SiPLS-PCA/LDA models. Furthermore, spectral data from VNIRS-CSA were the best for LDA with a high prediction value of 0.925 after standard normal variate (SNV) processing and variable selection by SiPLS. The research demonstrated that VNIRS-CSA is a quick, accurate, and non-destructive method for monitoring the storage time of rice. The strategy also has the potential for volatile organic components analysis.

HighlightsWhile freshness is a critical value of food quality, its assessment requires complex methods, which are costly and time-consuming.In this work, it is demonstrated that spectral responses obtained from a portable VIS/NIR imaging spectrometer can be used to predict food freshness using a CNN-based machine learning algorithm.In the food industry, the method can assess real-time food freshness nondestructively and cost-effectively.Abstract. There has been strong demand for the development of accurate but simple methods to assess the freshness of foods. In this study, a system is proposed to determine the freshness of fish by analyzing the spectral response with a portable visible/near-infrared (VIS/NIR) imaging spectrometer and a convolution neural network (CNN) machine learning algorithm. Spectral response data from salmon and tuna, which were incubated at 25°C, were obtained every minute for 30 h and were categorized into three stages (fresh, likely spoiled, or spoiled) based on the time and pH. Using the obtained spectral data, a CNN-based machine learning algorithm was built to evaluate the freshness of the experimental samples. The accuracy of the spectral data in predicting the freshness was ~84% for salmon and ~88% for tuna. Keywords: CNN, Fish, Freshness, pH, Spectral data, VIS/NIR.

Miso, the Japanese fermented soybean paste, an essential seasoning for preparing washoku, is gaining increased international recognition. Miso quality is normally evaluated by a panel of sensory as...

Soluble solids content (SSC) is one of the most important quality attributes affecting the taste and maturity of fresh fruit. In this study, with the cerasus humilis fruit as the research object, a prediction model of soluble solid content (SSC) in cerasus humilis (CH) is established based on visible / near-infrared spectroscopy to explore a nondestructive testing method of the interior quality of CH. The visible / near-infrared spectral info (350-2500nm) of 160 CHs was collected to extract the reflection spectrum, establishing the linear model (PLSR) and non-linear model (LS-SVM) of CH’s spectral info and SSC. The prediction performance and stability of the model were justified using several statistical indicators namely correlation coefficient of the prediction set (Rp), the root mean square error of the prediction set (RMSEP), and the residual predictive deviation (RPD) index. Results showed that multiplicative scatter correction (MSC) was proved to be the best preprocessing method, UVE-CARS was the optimal method of dimension reduction, the quantities of characteristic wavelengths was 10 and the optimal model was UVE-CARS-PLSR, in which Rc is 0.8995, Rp is 0.8579, RMSEC is 0.8897, RMSEP is 0.9059, and RPD is 1.8766, indicating that the redundant data of the original spectrum can be reduced, the wavelength dimensions can be reduced, valid info can be retained and data processing can be simplified as UVE-CARS extracts characteristic wavelengths. Reference and theoretical basis are provided in this research for future research and development of portable detector and online sorting detection of CH internal quality.

In this study, visible/near-infrared (Vis/NIR) hyperspectral imaging was used for the nondestructive detection of storage time of strawberries. Storage time was calculated immediately after freshly picking. Support vector machine (SVM) with multiplicative scatter correction can differentiate strawberries of different storage time with an accuracy of 100%. Then, the model developed by partial least square regression with full-range spectra was used to predict the storage time of strawberries with a determination coefficient of prediction (Rp2) of 0.9999 and root-mean-square error of prediction (RMSEP) of 0.0721, and deviation was small at different periods. With the spectra of 10 important wavelengths obtained by uninformative variable elimination, the SVM model obtained relatively acceptable results with Rp2 of 0.9943 and RMSEP of 1.3213. The prediction experiments for the separately picked strawberry samples also got the similar results. Finally, distribution maps of storage time generated based on the pixel-wise spectra and established model clearly show the quality transformation of the strawberries.

Early Access; International audience; The biological processes of interest to agro-industry involve numerous bacterial species. Lactic acid bacteria produce metabolites capable of fermenting food products and modifying their organoleptic properties, and plant-growth-promoting rhizobacteria can act as biofertilizers, biostimulants, or biocontrol agents in agriculture. The protocol of conventional techniques for bacterial identification, currently based on genotyping and phenotyping, require specific sample preparation and destruction. The work presented herein details a method for rapid identification of lactic acid bacteria and rhizobacteria at the genus and species level. To develop the method, bacteria were inoculated on an agar medium and analyzed by near infrared (NIR) and ultraviolet-visible-NIR (UV-Vis-NIR) spectroscopy. Artificial neural network models applied to the UV-Vis-NIR spectra correctly identified the genus (species) of 70% (63%) of the lactic acid bacteria and 67% of the rhizobacteria on an independent prediction set of unknown bacterial strains. These results demonstrate the potential of UV-Vis-NIR spectroscopy to identify bacteria directly on agar plates.

To realise accurate and nondestructive detection on moisture content of maize seed based on visible/near‐infrared (Vis/NIR) and near‐infrared (NIR) hyperspectral imaging technology, the hyperspectral images on two sides (embryo and endosperm sides) of each maize seed of four varieties were collected. The effects of average spectra extraction regions, that is centroid region and whole seed region, and different spectral preprocessing methods, were investigated. Uninformative variable elimination (UVE) was used to extract the feature wavelengths, and the partial least squares regression (PLSR) prediction models were established. The results showed that extracting the average spectra from the centroid region did better than from the whole seed region, and S‐G smoothing was prior to other preprocessing methods. The PLSR models established with NIR spectra had better performance than that with Vis/NIR spectra. The model developed for a single variety was superior to that for all varieties together.

Unripe banana flour is a premium nutritious product with a potential to curb degenerative diseases through resistant starch and gluten free traits, however, with scant techniques to monitor adulteration practices. The objective of the present study was to determine the efficacy of visible–near infrared spectroscopy (Vis–NIR) spectroscopy (Vis–NIRS) in the detection and quantification of unripe banana flour adulteration with wheat flour. To do this, simulated adulteration of a composite banana flour was performed with different levels of wheat flour, in intervals of 20 g kg(−1), ranging from 0 to 800 g kg(−1). Each level was acquired in duplicate giving a total of 82 samples. Vis–NIR spectral data was acquired using a portable F-750 spectrometer in the range 447–1005 nm. Spectral data was analysed chemometrically using principle components analysis and partial least squares regression (PLSR), with 41 samples used as a calibration set and 41 for validation. The first two principal components accounted for 95% of spectral data variation, revealing five distinct clusters related to 0 g kg(−1), 20–200 g kg(−1), 220–400 g kg(−1), 420–600 g kg(−1) and 620–800 g kg(−1) adulterated samples. The best PLSR model to predict wheat flour adulteration degrees in unripe banana flour was obtained using 2nd derivative Savitzky–Golay (19-point smoothing, 2nd order polynomial), showing the highest R(c)(2) (0.991); R(p)(2) (0.993); RPD (12.021) and the lowest RMSEC (2.226 g kg(−1)) and RMSEP (1.993 g kg(−1)) values. The obtained Vis–NIRS PLSR models therefore demonstrated the technology novelty in monitoring unripe banana flour quality by the processing industries and in retail markets during product verification.

Bacterial biofilms constitute a major source of sanitary problems and economic losses in the food industry. Indeed, biofilm removal may require intense mechanical cleaning procedures or very high concentrations of disinfectants or both, which can be damaging to the environment and human health. This study assessed the efficacy of a technique based on spectroscopy in the visible, near-infrared, and short-wavelength infrared range for the quick detection of biofilms formed on polystyrene by the pathogenic bacterium Staphylococcus aureus. To do that, biofilms corresponding to three S. aureus strains, which differed in biofilm-forming ability and composition of the extracellular matrix, were allowed to develop for 5 or 24 h, representing an active formation stage and mature biofilms, respectively. Spectral analysis of the samples, corresponding to three biological replicates of each condition, was then performed by using a portable device. The results of these experiments showed that partial least-squares discriminant analysis of the spectral profile could discriminate between surfaces containing attached bacterial biomass and noninoculated ones. In this model, the two first principal components accounted for 39 and 19% of the variance and the estimated error rate stabilized after four components. Cross-validation accuracy of this assessment was 100%. This work lays the foundation for subsequent development of a spectroscopy-based protocol that allows biofilm detection on food industrial surfaces., This work was funded by the following grants: Program of Science, Technology and Innovation 2013–2017 (AGL2015-65673-R), Ministry of Economy, Industry and Competitiveness, Spain (EU Animal Health and Welfare [ANIHWA] ERA-Net; BLAAT ID: 67/PCIN-2017-001), European Regional Development Fund (ERDF) EU funds, Principado de Asturias, Spain (Proyecto Intramural CSIC 201770E016, GRUPIN14-139 and IDI/2018/000119), Ministerio de Economía y Competitividad, Spain (BIO2017-88147-R), and Programa de Apoyo y Promoción de la Investigación, University of Oviedo, Spain (PAPI 2018)

Background Rice adulteration in the food industry that infringes on the interests of consumers is considered very serious. To realize the rapid and precise quantitation of adulterated rice, a visible near infrared (VNIR) hyperspectral imaging system (380-1000 nm) was developed in the present study. A Savitsky-Golay first derivative (SG1) transform was utilized to eliminate the constant spectral baseline offset. Then, the adulterated levels of rice samples were quantified by partial least squares regression (PLSR). Results A SG1-PLSR model based on full-wavelength was attained with a coefficient of determination of prediction set (RP ) of 0.9909, root-mean-square error of prediction set (RMSEP ) of 0.0447 g kg-1 and residual predictive deviation (RPDP ) of 11.28. Furthermore, fifteen important wavelengths were selected based on the weighted regression coefficients (BW ) and a simplified model (PLSR-15) was established with RP of 0.9769, RMSEP of 0.0708 g kg-1 and RPDP of 3.49. Finally, two visualization maps produced by applying the optimal models (SG1-PLSR and PLSR-15) were used to visualize the adulterated levels of rice. Conclusion These results demonstrate that VNIR hyperspectral imaging system is an effective tool for rapidly quantifying and visualizing the adulterated levels of rice. © 2019 Society of Chemical Industry.

Salmonella is an organism of importance to the poultry industry with increasingly stringent government regulatory standards. Real-time polymerase chain reaction (RT-PCR) and plating procedures on nutrient enriched growth media have been the standard detection methods of Salmonella from broiler chicken carcasses for years. These methods are proven, but offer disadvantages in the amount of time or reoccurring sample cost. Here, we propose the use of a hyperspectral microscope imaging system (HMI) for comparison to standard detection methods. Broiler chicken carcasses were rinsed and plated on Salmonella selective agar. Colonies from plates were picked and RT-PCR was used as a confirmation test to verify plating results, while HMI was collected from the same colonies. Spectral signatures of cells were extracted between 450 and 800 nm from HMI collected with 100x objective. A quadratic discriminant analysis (QDA) was used to classify cells as either Salmonella positive or negative (n = 341). Spectra preprocessing minimized the influence of cellular shape on the spectra, increasing the initial classification accuracy of 81.8–98.5%, yielding a sensitivity of 1.0, and a specificity of 0.963. Results showed the potential as an initial investigation of HMI as a microbial confirmation tool, compared to RT-PCR.

Current methods for detecting aflatoxin contamination of agricultural and food commodities are generally based on wet chemical analyses, which are time-consuming, destructive to test samples, and require skilled personnel to perform, making them impossible for large-scale nondestructive screening and on-site detection. In this study, we utilized visible–near-infrared (Vis-NIR) spectroscopy over the spectral range of 400–2500 nm to detect contamination of commercial, shelled peanut kernels (runner type) with the predominant aflatoxin B1 (AFB1). The artificially contaminated samples were prepared by dropping known amounts of aflatoxin standard dissolved in 50:50 (v/v) methanol/water onto peanut kernel surface to achieve different contamination levels. The partial least squares discriminant analysis (PLS-DA) models established using the full spectra over different ranges achieved good prediction results. The best overall accuracy of 88.57% and 92.86% were obtained using the full spectra when taking 20 and 100 parts per billion (ppb), respectively, as the classification threshold. The random frog (RF) algorithm was used to find the optimal characteristic wavelengths for identifying the surface AFB1-contamination of peanut kernels. Using the optimal spectral variables determined by the RF algorithm, the simplified RF-PLS-DA classification models were established. The better RF-PLS-DA models attained the overall accuracies of 90.00% and 94.29% with the 20 ppb and 100 ppb thresholds, respectively, which were improved compared to using the full spectral variables. Compared to using the full spectral variables, the employed spectral variables of the simplified RF-PLS-DA models were decreased by at least 94.82%. The present study demonstrated that the Vis-NIR spectroscopic technique combined with appropriate chemometric methods could be useful in identifying AFB1 contamination of peanut kernels.

Pistachios are susceptible to aflatoxin contamination because of their rich nutrient content. Hyperspectral imaging (HSI), a new method for collecting spectral and image information, has been successfully employed in contamination research to classify staple agricultural products, such as maize, that are contaminated with aflatoxins. However, only a few studies have been conducted on the nondestructive discrimination among contaminated nuts using HSI for both qualitative and quantitative purposes. Thus, the feasibility of directly detecting aflatoxin B1 (AFB1) in individual pistachio kernels using visible/near-infrared HSI (VNIR HSI) was explored in this study. A total of 300 pistachio kernels were randomly selected to prepare target samples that were artificially contaminated with 5, 10, 20, 30, or 50 ppb (parts per billion) of AFB1. Principal component analysis (PCA) showed an overall separation trend between the control and all contaminated kernels. Accuracies greater than 90.0% were obtained by linear discriminant analysis (LDA) for samples that were artificially contaminated with different concentrations of AFB1 based on spectra at 694 to 988 nm that had been preprocessed with standard normal variate (SNV) and Savitzky-Golay (SG) smoothing. The correlation coefficients of calibration and validation (rc and rv) from stepwise multiple linear regression (SMLR) models were all >0.9100. Moreover, five key wavelengths (708, 771, 892, 915, and 941 nm) closely associated with AFB1 contamination were identified using principal component spectra analysis. Generally, the results indicated that VNIR HSI could be employed for preliminary screening of pistachio kernels that were artificially contaminated with AFB1, even at the 5 ppb level. However, the quantitative prediction of the specific AFB1 concentration needed to be further improved. Keywords: Aflatoxins, Detection analysis, Hyperspectral information, Pistachios, Visible/near-infrared.

Visible/Near-infrared (Vis/NIR) spectroscopy at a range of 450–1000 nm was used to predict the values of three qualitative variables (starch, reducing sugar, and moisture content) on 200 potato tubers from 2 potato genotypes (‘Agria’ and ‘Clone 397009–8’) stored in both traditional and cold storages. After spectroscopy measurements, these variables were measured using reference methods. Then, Partial Least Square (PLS) models were developed. To evaluate developed models, Root Mean Square Error of calibration and cross validation (RMSEC and RMSECV), as well as coefficient of determination for calibration and cross validation (R 2 C and R 2 CV), and Residual Predictive Deviation (RPD) were used. The best prediction belonged to reducing sugar with statistical values of R 2 C = 0.99, R 2 CV = 0.98, RMSEC = 0.029, RMSECV = 0.037, and RPD = 7.57 in ‘Clone’ genotype stored under cold storage. The weakest prediction was related to moisture content with statistical values of R 2 C = 0.93, R 2 CV = 0.92, RMSEC = 0.268, RMSECV = 0.279, and RPD = 6.45 in stored ‘Clone’ genotypes under cold storage. Results of the study showed that, Vis/NIR spectroscopy as a non-destructive, fast, and reliable technique can be used for prediction of inner compositions of stored potatoes.

With increasing consumer interest in healthy food formulations made from pulse flours, the research gap in characterizing them based on their pulse type and milling technique has come to limelight. To this end, the feasibility of using hyperspectral imaging in the visible near infrared (Vis-NIR) (400–1000 nm) and short wave infrared (SWIR) (1000–2500 nm) regions to classify pulse flours (viz. chickpea, yellow pea, navy bean and green lentil) based on the pulse type and milling methods was investigated. Unsupervised and supervised classification models were developed using unsupervised principal component analysis (PCA) and supervised partial least squares discriminant analysis (PLS-DA). Supervised classification for pulse flour type demonstrated 100% accuracy in the Vis-NIR wavelength range of 530–700 nm, which primarily exploited the color attributes of the flour samples. For milling method based classification, PLS-DA models developed using SWIR regions of 1370–1500 nm and 1700–2000 nm played a significant role in discriminating flour samples yielding 95% classification accuracy. These regions are associated with the O–H and N–H overtones of the proteins found in flour samples. Conclusively, hyperspectral imaging in the range of 400–2500 nm combined with multivariate data classification methods can reliably be used by the food industry for characterization of pulse flours.

A visible/near-infrared (VNIR) hyperspectral imaging (HSI) system (400–1000 nm) was used to assess the feasibility of detecting aflatoxin B1 (AFB1) on surfaces of 600 kernels of four maize varieties from different regions of the U.S.A. i.e. Georgia, Illinois, Indiana and Nebraska. For each variety, four AFB1 solutions (10, 20, 100 and 500 ppb) were artificially applied on kernel surfaces. Similarly, a control group was generated from 30 kernels of each variety treated with a solution of methanol. Principal component analysis (PCA) was used to reduce dimensionality of the HSI data followed by the application of factorial discriminant analysis (FDA) on the principal component variables. PCA results showed a pattern of separation between uncontaminated and contaminated kernels for all varieties except for Indiana and pooled samples. FDA showed the ability to predict AFB1 contamination of each variety with over 96% validation accuracy while prediction for AFB1 contamination group membership of pooled samples reached 98% accuracy in validation. Variation in the spectra of AFB1 contaminated kernels could have caused the variation in the predicted AFB1 contamination group membership. The PCA and FDA models where influenced by the chemical information from C H, N H and O H bonds of VNIR spectral regions. This study presents the potential of using VNIR hyperspectral imaging in direct AFB1 contamination classification studies of maize kernels of different varieties. The study further suggests that varietal differences of maize kernels may have no influence on AFB1 contamination classification.

Fungal diseases are among the most common problems that affect tomato fruits, especially after harvesting. The early detection of these diseases may save 60% of the losses. Thus, this study aimed to detect the gray mold disease at an early stage on tomato fruits using the non-destructive visible/near-infrared (VIS/NIR) technique. Botrytis. cinerea fungus was isolated and identified using two specific primer sets (C729+/- and BC108+/BC563-). Three tomato varieties were used for this study (i.e. Harver, Izmer and Ekram), 30 samples from each variety, 20 samples were injected with the B. cinerea and 10 samples were left as control/healthy samples. Samples were examined using the VIS/NIR spectroscopy with a range of 550-1100 nm for three days. Thereafter, the data were analyzed using principal component analysis (PCA), where the results showed the ability of VIS/NIR to detect the infected samples on the second day of the injection, and before the symptoms appear on the samples. Two PCs explained 99% and 100% of VIS/NIR variance, respectively.

This study aimed to assess the feasibility of identifying multiple chemical constituents in matcha using visible-near infrared hyperspectral imaging (VNIR-HSI) technology. Regions of interest (ROIs) were first defined in order to calculate the representative mean spectrum of each sample. Subsequently, the standard normal variate (SNV) method was applied to correct the characteristic spectra. Competitive adaptive reweighted sampling (CARS) and bootstrapping soft shrinkage (BOSS) were used to optimize the models. They were built based on partial least squares (PLS), creating two models referred to as CARS-PLS and BOSS-PLS. The BOSS-PLS models achieved best predictive accuracy, with coefficients of determination predicted to be 0.8077 for caffeine, 0.7098 for tea polyphenols (TPs), 0.7942 for free amino acids (FAAs), 0.8314 for the ratio of TPs to FAAs, and 0.8473 for chlorophyll. These findings highlight the potential of VNIR-HSI technology as a rapid and nondestructive alternative for simultaneous quantification of chemical constituents in matcha.

BACKGROUND The market for fruits and vegetables is mainly controlled by the mass distribution channel (MDC). MDC buyers do not have useful instruments to rapidly evaluate the quality of the products. Decisions by the buyers are driven primarily by pricing strategies rather than product quality. Simple, rapid and easy-to-use methods for objectively evaluating the quality of postharvest products are needed. The aim of this study was to use visible and near-infrared (vis/NIR) spectroscopy to estimate some qualitative parameters of two low-price products (carrots and tomatoes) of various brands, and to evaluate the applicability of this technique for use in stores. RESULTS A non-destructive optical system (vis/NIR spectrophotometer with a reflection probe, spectral range 450–1650 nm) was tested. The differences in quality among carrots and tomatoes purchased from 13 stores on various dates were examined. The reference quality parameters (firmness, water content, soluble solids content, pH and colour) were correlated to the spectral readings. The models derived from the optical data gave positive results, in particular for the prediction of the soluble solids content and the colour, with better results for tomatoes than for carrots. CONCLUSION The application of optical techniques may help MDC buyers to monitor the quality of postharvest products, leading to an effective optimization of the entire supply chain.

In this study, visible and near-infrared hyperspectral imaging (HSI) technique combined with deep learning algorithm was investigated for discriminating the freshness of shrimp during cold storage. Shrimps were labeled into two freshness grades (fresh and stale) according to their total volatile basic nitrogen contents. Spectral features were extracted from the HSI data by stacked auto-encoders (SAEs)-based deep learning algorithm and then used to classify the freshness grade of shrimp by a logistic regression (LR)-based deep learning algorithm. The results demonstrated that the SAEs–LR achieved satisfactory total classification accuracy of 96.55 and 93.97% for freshness grade of shrimp in calibration (116 samples) and prediction (116 samples) sets, respectively. An image processing algorithm was also developed for visualizing the classification map of freshness grade. Results confirmed the possibility of rapid and nondestructive detecting freshness grade of shrimp by the combination of hyperspectral imaging technique and deep learning algorithm. The SAEs–LR method adds a new tool for the multivariate analysis of hyperspectral image for shrimp quality inspections.

Background: Texture is a major quality parameter for the acceptability of canned whole beans. Prior knowledge of this quality trait before processing would be useful to guide variety development by bean breeders and optimize handling protocols by processors. The objective of this study was to evaluate and compare the predictive power of visible and near infrared reflectance spectroscopy (visible/NIRS, 400-2498 nm) and hyperspectral imaging (HYPERS, 400-1000 nm) techniques for predicting texture of canned black beans from intact dry seeds. Black beans were grown in Michigan (USA) over three field seasons. The samples exhibited phenotypic variability for canned bean texture due to genetic variability and processing practice. Spectral preprocessing methods (i.e. smoothing, first and second derivatives, continuous wavelet transform, and two-band ratios), coupled with a feature selection method, were tested for optimizing the prediction accuracy in both techniques based on partial least squares regression (PLSR) models.; Results: Visible/NIRS and HYPERS were effective in predicting texture of canned beans using intact dry seeds, as indicated by their correlation coefficients for prediction (Rpred ) and standard errors of prediction (SEP). Visible/NIRS was superior (Rpred = 0.546-0.923, SEP = 7.5-1.9 kg 100 g-1 ) to HYPERS (Rpred = 0.401-0.883, SEP = 7.6-2.4 kg 100 g-1 ), which is likely due to the wider wavelength range collected in visible/NIRS. However, a significant improvement was reached in both techniques when the two-band ratios preprocessing method was applied to the data, reducing SEP by at least 10.4% and 16.2% for visible/NIRS and HYPERS, respectively. Moreover, results from using the combination of the three-season data sets based on the two-band ratios showed that visible/NIRS (Rpred = 0.886, SEP = 4.0 kg 100 g-1 ) and HYPERS (Rpred = 0.844, SEP = 4.6 kg 100 g-1 ) models were consistently successful in predicting texture over a wide range of measurements.; Conclusion: Visible/NIRS and HYPERS have great potential for predicting the texture of canned beans; the robustness of the models is impacted by genotypic diversity, planting year and phenotypic variability for canned bean texture used for model building, and hence, robust models can be built based on data sets with high phenotypic diversity in textural properties, and periodically maintained and updated with new data. © 2017 Society of Chemical Industry.; © 2017 Society of Chemical Industry.

A partial least squares regression (PLSR) model to map internal soluble solids content (SSC) of apples using visible/near-infrared (VNIR) hyperspectral imaging was developed. The reflectance spectra of sliced apples were extracted from hyperspectral absorbance images obtained in the 400–1000 nm range. Prediction models for SSC mapping were developed for three different measurement/sampling designs that varied in the number and size of the regions of interest (ROIs) used for apple SSC measurement and spectral averaging. Case 1 used 29 small ROIs per apple, Case II used 9 moderate-size ROIs per apple, and Case III used 5 large ROIs per apple. The optimal pre-treatment of the spectra extracted from the hyperspectral images was investigated to enhance the performance of the prediction models. The coefficients of determination and root mean square errors of the best-performing models were, respectively, 0.802 and ±0.674 °Brix for Case I, 0.871 and ±0.524 °Brix for Case II, and 0.876 and ±0.514 °Brix for Case III. The accuracy of the PLSR models was enhanced by using the spectra and SSC measured/averaged from the fewer but larger areas of the apples rather than from more numerous but smaller areas. PLS images of SSC showed the predicted internal distribution of SSC within the apples. The overall results demonstrate that hyperspectral absorbance imaging techniques may be useful for mapping internal soluble solids content of apples.

The potential of visible&ndash, near-infrared (Vis&ndash, NIR) spectroscopy to predict physico-chemical quality traits in 368 samples of bovine musculus longissimus thoracis et lumborum (LTL) was evaluated. A fibre-optic probe was applied on the exposed surface of the bovine carcass for the collection of spectra, including the neck and rump (1 h and 2 h post-mortem and after quartering, i.e., 24 h and 25 h post-mortem) and the boned-out LTL muscle (48 h and 49 h post-mortem). In parallel, reference analysis for physico-chemical parameters of beef quality including ultimate pH, colour (L, a*, b*), cook loss and drip loss was conducted using standard laboratory methods. Partial least-squares (PLS) regression models were used to correlate the spectral information with reference quality parameters of beef muscle. Different mathematical pre-treatments and their combinations were applied to improve the model accuracy, which was evaluated on the basis of the coefficient of determination of calibration (R2C) and cross-validation (R2CV) and root-mean-square error of calibration (RMSEC) and cross-validation (RMSECV). Reliable cross-validation models were achieved for ultimate pH (R2CV: 0.91 (quartering, 24 h) and R2CV: 0.96 (LTL muscle, 48 h)) and drip loss (R2CV: 0.82 (quartering, 24 h) and R2CV: 0.99 (LTL muscle, 48 h)) with lower RMSECV values. The results show the potential of Vis&ndash, NIR spectroscopy for online prediction of certain quality parameters of beef over different time periods.

The appearance of white striations over breast meat is an emerging and growing problem. The main purpose of this study was to employ the reflectance of visible-near infrared (VIS/NIR) spectroscopy to differentiate between normal and white striped turkey breasts. Accordingly, 34 turkey breast fillets were selected representing a different level of white striping (WS) defects (normal, moderate and severe). The findings of VIS/NIR were analyzed by principal component (PC1) analysis (PCA). It was found that the first PC1 for VIS, NIR and VIS/NIR region explained 98%, 97%, and 96% of the total variation, respectively. PCA showed high performance to differentiate normal meat from abnormal meat (moderate and severe WS). In conclusion, the results of this research showed that VIS/NIR spectroscopy was satisfactory to differentiate normal from severe WS turkey fillets by using several quality traits.

International audience; The aim of this study was to compare visible-near-infrared spectroscopy (VIS/NIRS) models developed from fresh or freeze-dried samples for predicting the fatty acid (FA) composition of beef samples. The hypothesis tested is that the removal of water from samples could improve the VIS/NIRS model performance. A total of 454 beef samples obtained from different bovine muscles were used. No significant differences were found in the performance of VIS/NIRS models developed from fresh or freeze-dried samples for predicting both major individual FAs and families of FAs and for some FAs (16:0, 18:0, 18:1 n-9, 18:2 n-6, 20:4 n-6, 22:5 n-3, 22:6 n-3, saturated, mono-unsaturated FA, and total n-3 long chain poly-unsaturated FAs (PUFA)). In contrast, the standard error of predictions for total PUFAs, total n-3 PUFAs, total conjugated linoleic acid, 20:5 n-3, and 18:3 n-3 were improved (by 21% on average; P < .05) in freeze-dried samples compared with fresh samples.

The particular characteristics of some of the Ethiopian mustard accessions available from seed banks could be used to increase the production and the diversity of products available to consumers and to improve their general quality. The objectives of this study were to determine the genetic variability for agro-morphological (days to first flowering: DFF and leaf pubescence: LP) and nutritional traits (total phenolic content: TPC) among accessions, and to evaluate the potential of near-infrared spectroscopy (NIRS) to predict these traits in Ethiopian mustard leaves. A great variation was found for the traits evaluated. The reference values were regressed against different spectral transformations by modified partial least-squares (MPLS) regression. The coefficients of determination in cross-validation (R2cv) shown by the equations for DFF, LP and TPC were 0.95, 0.63 and 0.99, respectively. The standard deviation to standard error of cross-validation ratio (RPD), were for these traits, as follows: DFF: 4.52, LP: 1.53 and, TPC: 24.50. These results show that the equations developed for DFF and TPC in Ethiopian mustard, can be predicted with sufficient accuracy for screening purposes and quality control, respectively. In addition, the LP equation can be used to identify those samples with &ldquo, low&rdquo, &ldquo, medium&rdquo, and &ldquo, high&rdquo, groups. From the study of the mean and deviation standard spectra, and regression vectors of MPLS models it can be concluded that some major cell components, highly participated in modelling the equations for these traits.

In this study, the potential of visible and near infrared spectroscopy was investigated to classify the maturity stage and to predict the quality attributes of pomegranate variety “Ashraf” such as total soluble solids content, pH, and titratable acidity during four distinct maturity stages between 88 and 143 days after full bloom. Principal component analysis was used to distinguish among different maturities. The prediction models of internal quality attributes of the pomegranate were developed by partial least squares regression. The transmission spectra of pomegranate were obtained in the wavelength range from 400 to 1100 nm. In this research several preprocessing methods were utilized including centering, smoothing (Savitzky–Golay algorithm, median filter), normalization (multiplicative scatter correction and standard normal variate) and differentiation (first derivative and second derivative). It concluded that different preprocessing techniques had effects on the classification performance o...

Background: There is a need to minimize economic damage by sorting infected seeds from healthy seeds before seeding. However, current methods of detecting infected seeds, such as seedling grow-out, enzyme-linked immunosorbent assays, the polymerase chain reaction (PCR) and the real-time PCR have a critical drawbacks in that they are time-consuming, labor-intensive and destructive procedures. The present study aimed to evaluate the potential of visible/near-infrared (Vis/NIR) hyperspectral imaging system for detecting bacteria-infected watermelon seeds.; Results: A hyperspectral Vis/NIR reflectance imaging system (spectral region of 400-1000 nm) was constructed to obtain hyperspectral reflectance images for 336 bacteria-infected watermelon seeds, which were then subjected to partial least square discriminant analysis (PLS-DA) and a least-squares support vector machine (LS-SVM) to classify bacteria-infected watermelon seeds from healthy watermelon seeds. The developed system detected bacteria-infected watermelon seeds with an accuracy > 90% (PLS-DA: 91.7%, LS-SVM: 90.5%), suggesting that the Vis/NIR hyperspectral imaging system is effective for quarantining bacteria-infected watermelon seeds.; Conclusion: The results of the present study show that it is possible to use the Vis/NIR hyperspectral imaging system for detecting bacteria-infected watermelon seeds. © 2016 Society of Chemical Industry.; © 2016 Society of Chemical Industry.

The aim of this work was to investigate the influence of packaging on measurements of visible-near infrared and near infrared spectroscopy during the shelf life of fresh-cut Valerianella leaves and Golden Delicious apple slices. Principal component analysis and partial least square discriminant analysis were applied on spectral data to reach the objectives. Results on Golden Delicious could be considered satisfactory even if the classification performances were lower than those obtained for Valerianella; however, overall collected positive predictive values ranged from 86.7% to 100.0%. The presence of packaging had only a slight effect on classification capability. These findings would be useful in the development of freshness decay assessment tools using these optical techniques along products' post-packaging chain.

This paper described a rapid and non-destructive method based on visible near-infrared (Vis-NIR) hyperspectral imaging system (400–1000 nm) for detection adulteration with duck meat in minced lamb. The multiple average of the reference spectral and a predicted relative spatial distribution coefficient were applied in this study to reduce the noise of the spectra. The PLSR model with selected wavelengths achieved better results than others with determination of coefficients (R2P) of 0.98, and standard error of prediction (RMSEP) of 2.51%. And the prediction map of the duck minced in lamb meat was generated by applying the prediction model. The results of this study indicate the great potential of the hyperspectral technology applying to rapidly and accurately detect the meat adulteration in minced lamb meat.

The feasibility of using visible/near-infrared (Vis/NIR) spectroscopy was assessed for non-destructive detection of diazinon residues in intact cucumbers. Vis/NIR spectra of diazinon solution and cucumber samples without and with different concentrations of diazinon residue were analysed at the range of 450-1000 nm. Partial least squares-discriminant analysis (PLS-DA) models were developed based on different spectral pre-processing techniques to classify cucumbers with contents of diazinon below and above the MRL as safe and unsafe samples, respectively. The best model was obtained using a first-derivative method with the lowest standard error of cross-validation (SECV = 0.366). Moreover, total percentages of correctly classified samples in calibration and prediction sets were 97.5% and 92.31%, respectively. It was concluded that Vis/NIR spectroscopy could be an appropriate, fast and non-destructive technology for safety control of intact cucumbers by the absence/presence of diazinon residues.

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.

For the first time, a visible near-infrared (Vis-NIR) hyperspectral imaging system (400–1000 nm) was investigated for rapid and non-destructive detection of adulteration in minced beef meat. Minced beef meat samples were adulterated with horsemeat at levels ranging from 2 to 50 % (w/w), at approximately 2 % increments. Calibration model was developed and optimized using partial least-squares regression (PLSR) with internal full cross-validation and then validated by external validation using an independent validation set. Several spectral pre-treatment techniques including derivatives, standard normal variate (SNV), and multiplicative scatter correction (MSC) were applied to examine the influence of spectral variations for predicting adulteration in minced beef. The established PLSR models based on raw spectra had coefficients of determination (R2) of 0.99, 0.99, and 0.98, and standard errors of 1.14, 1.56, and 2.23 % for calibration, cross-validation, and prediction, respectively. Four important wavelengths (515, 595, 650, and 880 nm) were selected using regression coefficients resulting from the best PLSR model. By using these important wavelengths, an image processing algorithm was developed to predict the adulteration level in each pixel in whole surface of the samples. The results demonstrate that hyperspectral imaging coupled with multivariate analysis can be successfully applied as a rapid screening technique for adulterate detection in minced meat.

BACKGROUND Wineries need new practical and quick instruments, non-destructive and able to quantitatively evaluate during withering the parameters that impact product quality. The aim of the work was to test an optical portable system (visible near-infrared (NIR) spectrophotometer) in a wavelength range of 400–1000 nm for the prediction of quality parameters of grape berries during withering. RESULTS A total of 300 red grape samples (Vitis vinifera L., Corvina cultivar) harvested in vintage year 2012 from the Valpolicella area (Verona, Italy) were analyzed. Qualitative (principal component analysis, PCA) and quantitative (partial least squares regression algorithm, PLS) evaluations were performed on grape spectra. PCA showed a clear sample grouping for the different withering stages. PLS models gave encouraging predictive capabilities for soluble solids content (R2val = 0.62 and ratio performance deviation, RPD = 1.87) and firmness (R2val = 0.56 and RPD = 1.79). CONCLUSION The work demonstrated the applicability of visible NIR spectroscopy as a rapid technique for the analysis of grape quality directly in barns, during withering. The sector could be provided with simple and inexpensive optical systems that could be used to monitor the withering degree of grape for better management of the wine production process. © 2014 Society of Chemical Industry

In this study visible/near-infrared spectroscopy (Vis/NIRS) was evaluated to rapidly classify intact chicken breast fillets. Five principal components (PC) were extracted from reference quality traits (L∗, pH, drip loss, expressible fluid, and salt-induced water gain). A quality grades classification method by PC1 score was proposed. With this method, 150 chicken fillets were properly classified into three quality grades, i.e., DFD (dark, firm and dry), normal, and PSE (pale, soft and exudative). Furthermore, PC1 score could be predicted using partial least squares regression (PLSR) model based on Vis/NIRS (R2p = 0.78, RPD = 1.9), without the measurement of any quality traits. Thresholds of PC1 classification method were applied to classify the predicted PC1 score values of each fillet into three quality grades. The classification accuracy of calibration and prediction set were 85% and 80%, respectively. Results revealed that PC1 score classification method is feasible, and with Vis/NIRS, this method could be rapidly implemented.

The main objective of this study was to investigate the use of spectroscopic systems in the range of 400-1000nm (visible/near-infrared or Vis-NIR) and 900-1700nm (NIR) to assess and estimate plant and animal proteins as potential adulterants in minced beef and pork. Multiple machine learning techniques were used for classification, adulterant prediction, and wavelength selection. Samples were first evaluated for the presence or absence of adulterants (6 classes), and secondly for adulterant type (6 classes) and level. Selected wavelengths models generally resulted in better classification and prediction outputs than full wavelengths. The first stage classification rates were 96% and 100% for pure/unadulterated and adulterated samples, respectively. Whereas, the second stage had classification rates of 69-100%. The optimal models for predicting adulterant levels yielded correlation coefficient, r of 0.78-0.86 and ratio of performance to deviation, RPD, of 1.19-1.98. The results from this study illustrate potential application of spectroscopic technology to rapidly and accurately detect adulterants in minced beef and pork.

International audience; The potential of multispectral VIS-NIR imaging to discriminate beef meat muscles in relation with their type and animal origin was examined in the present study. Two hundred and forty muscles of three types (Longissimus thoracis, Biceps femoris and Semimembranosus) were obtained from the carcasses of three types of animals, two late-maturing cattle types of animals (Limousin and Blond d'Aquitaine) that grow slowly and deposit more muscles and less fat, compared to one early-maturing cattle types of animals (Angus) which tends to have muscles richer in collagen and in intramuscular fat. Two hundred and forty cube images were collected with nineteen emitting LEDs (405 to 1050 nm) using the Videometer Lab2 device. The image cubes were processed in order to extract image mean spectra and image shape features from co-occurrence and difference of histogram matrices. The results of the PLSDA performed on image texture features and spectral data show a maximum ranging from 63.5 to 83% of good classification depending on the muscle and breed considered. This study demonstrated the promising potential of the VIS-NIR multispectral imager to characterize beef meat muscles based on muscle type and its animal origin.

Carrot is a root crop that has relatively high content of vitamin C and carotenoids. Carrot is usually consumed fresh, juiced or cooked. It is also a major source of industrial natural carotenoids, a precursor to vitamin A. Rapid determination of quality parameters indeed helps process engineers and storage managers to track the condition of root crops such as carrots during storage and/or before processing. The objectives of this study were to apply VIS-NIR spectroscopy to noninvasively assess and predict the various quality attributes of carrot (cv. Nectar) including color parameters (L*a*b*), moisture content (MC), total soluble solids (TSS), firmness, vitamin C content, and I²-carotene. Two spectroscopic sensors with different range were used (400-1000 nm and 900-1700 nm). Samples included whole roots and sliced discs. The best prediction models using partial least squares regression (PLSR) yielded correlation coefficient, r and ratio of performance to deviation or r(RPD) with values of 0.92 (2.46), 0.85(1.65), 0.94(2.85), 0.96(3.55), 1.00(17.92), 0.90(2.16), 0.98(4.73), and 0.97(3.88) for L*, a*, b*, firmness, MC, TSS, vitamin C, and I²-carotene, respectively. Generally, whole roots and VIS-NIR sensor have high prediction (r = 0.69 – 1.00). Results obtained in this research could be enhanced with more samples to develop a better and higher accuracy predictive multispectral model for carrot quality prediction with possible extension to other similar vegetables in post-harvest losses reduction.

Visible/near-infrared (Vis/NIR) spectroscopy was applied to determine the ripeness of melons (Cucumis melo L.). The total soluble solids (TSS) and flesh color distribution around the equator fracture surface was studied, and a high correlation between TSS and flesh color in the outer mesocarp region was observed. A new ripeness index, defined as the edible ratio (ER), was proposed. One hundred and twenty melons with similar sizes were selected as samples and divided into a calibration set and a prediction set. Spectra of melons were acquired separately under transmittance and interactance modes. A machine vision system was established to acquire images of the melon fracture surfaces to further calculate the ER values of the melons. Partial least squares (PLS) calibration models were developed to predict the melon ripeness indices, including TSS, surface color (SC), and ER value. The results were promising for TSS, Hunter a, and ER value. For Hunter b, the results were insufficient. In addition, by comparing the performances of the calibration models developed under different detection modes, it was concluded that the best calibration results for TSS and ER were achieved under the transmittance mode, while for Hunter a, the best result was obtained using interactance spectra.

VIS/NIR spectroscopic systems have been extensively and successfully applied on quality assurance for fruits, vegetables, and food products. This study uniquely focused on the efficacy of selected wavelengths on predicting glucose and sucrose for potato tubers of Frito Lay 1879 (FL) and Russet Norkotah (RN) cultivars, and in turn the potential of classification of potatoes based on sugar levels important to the frying industry. Whole tubers, as well as 12.7 mm slices, were scanned using a VIS/NIR interactance spectroscopic system (446–1,125 nm). Interval partial least squares, and genetic algorithm were utilized for wavelength selection. Partial least squares regression (PLSR), and artificial neural networks were applied for building prediction models. Results showed that R(RPD) [correlation coefficient (ratio of reference standard deviation to root mean square error of the model)] for prediction models of glucose were as high as 0.78 (1.61), and 0.95 (3.02) for FL, and RN for slice samples, and 0.81 (1.72), and 0.97 (3.89) for FL and RN respectively in the case of whole tubers. For sucrose models, R(RPD) values were 0.71 (1.43), and 0.78 (1.57) for FL and RN for slice samples, and 0.80 (1.64), and 0.94 (2.82) for whole tubers. Classification of potatoes based on sugar levels was conducted and training models were built and validated using linear discriminant analysis, K-nearest neighbor, partial least squares discriminant analysis, and classifier fusion. Classification errors of the testing set for whole tubers, based on glucose, were as low as 18 and 0 % for FL and RN. For sliced samples, the errors were 16 and 13 % for FL and RN. Generally, higher classification errors were obtained based on sucrose with values of whole tubers as low as 26 and 14 % for FL and RN, and for sliced samples the errors were 23 and 18 % which follows a similar trend as PLSR results. This study presents a potential of using selected wavelengths in the VIS/NIR interactance range of 446–1,125 nm to effectively predict sugars and classify potatoes based on thresholds that are crucial for the frying industry.

Near-infrared (NIR) transmittance spectroscopy, visible/NIR interactance spectroscopy, and visible/NIR hyperspectral imaging modes were used to determine the potential for rapid electronic estimation of glucose, sucrose, specific gravity, primordial leaf count, and soluble solids of (FL) (chipping) and Russet Norkotah (RN) (table) potato cultivars. Whole tubers and 0.5 inch (12.54 mm) slices were evaluated. Partial least squares regression (PLSR) was used to obtain the prediction models. Models for leaf counts and glucose were promising for all modes with the optimum model obtained for leaf count from interactance with sliced samples resulting in R (RPD) values of 0.95(3.29) for FL, and 0.90(2.19) for RN. For glucose, interactance also yielded the best model with R (RPD) values of 0.90(2.14) for FL, and 0.95(3.12) for RN. Models of other constituents were inconsistent between systems. Interactance mode with sliced samples showed best performance for soluble solids for FL with R (RPD) values of 0.55(1.18). Also, interactance mode demonstrated optimal performance for sucrose estimation with R (RPD) values of 0.81(1.63) for FL obtained from sliced samples, and 0.81(1.64) from whole tubers. Transmittance mode with sliced samples showed the best prediction models for RN with R (RPD) values of 0.63(1.30) for sucrose. Finally, specific gravity under interactance and transmittance modes with sliced samples for FL and RN respectively yielded R (RPD) values of 0.61(1.27), and 0.59(1.22).

Crucial physio-chemical changes occuring in eggs during storage after laying lead to loss of egg freshness. In this research, a new method for prediction of egg freshness using transmission visible near infrared spectroscopy was investigated. For this purpose 300 eggs were stored at two control conditions: refrigerator (4–5°C, 75%RH) and room (24–25°C, 40%RH) then by special egg holder, transmission spectroscopy was measured. For two eggs groups, 25 eggs in each group, in six days were tested by spectroscopy, after that Haugh unit and air cell height was measured directly. The non-destructive visible near infrared spectroscopy spectral measurements from 300 to 1100 nm (832 length of wave) were done as well as Haugh unit, air cell height for each egg and created the database for both environments. Finally a maximum likelihood latent root regression algorithm was developed to predict Haugh unit and air cell height by spectrum observation. The database was randomly divided into two parts. Training data, was ...

Reflectance Visible/near-infrared spectroscopy was used to analyze the effect of peel on the non-destructive soluble solids content prediction of oranges. The spectra were interpreted and partial least squares model was developed for both intact and peeled oranges at two wavelength ranges of 450–1000 and 700–1000 nm. Results showed that the wavelengths of the main absorbance peaks on the spectra of the intact and peeled samples were the same (p > 0.05). Moreover, the developed models yielded a good prediction accuracy of soluble solids content (2.5 0.05). It was also noted that the effect of peel chemical composition can be ignored in non-destructive spectroscopy of intact citrus for determination of soluble solids content.

Black bean (Phaseolus vulgaris L.) processing pre- sents unique challenges because of discoloration, breakage, development of undesirable textures, and off-flavors during canning and storage. These quality issues strongly affect processing standards and consumer acceptance for beans. In this research, visible and near-infrared (Vis/NIR) reflectance data for the spectral region of 400-2,500 nm were acquired from intact dry beans for predicting five canning quality traits, i.e., hydration coefficient (HC), visual appearance (APP) and color (COL), washed drained coefficient (WDC), and texture (TXT), using partial least squares regression (PLSR). A total of 471 bean samples harvested and canned in 2010, 2011, and 2012 were used for analysis. PLSR models based on the Vis/ NIR data showed low predictive performance, as measured by correlation coefficient for prediction (Rpred )f or APP (Rpred= 0.275-0.566) and TXT (Rpred=0.270-0.681), but better results for predicting HC (Rpred=0.517-0.810), WDC (Rpred=0.420- 0.796), and COL (Rpred

The freshness of rainbow trout is one of the most important quality parameters to attract customers. Common methods to detect fish freshness are usually subjective to the skill of a quality evaluator and are time consuming and destructive. Therefore, an automatic, nondestructive, accurate and quick method is needed. Hyperspectral imaging has demonstrated its efficiency in the meat and fish industries for quality control purposes. This method is nondestructive, fast and automatic. In this study, two setups for hyperspectral imaging named “Visible–Near Infrared” (Vis–NIR) and “Short Wave Infrared” (SWIR) are used to determine fish freshness. Eighty fresh rainbow trouts were divided into four batches which were separately preserved in ice for 1, 3, 5 and 7 days, respectively. Principle Component Analysis (PCA) and Partial Least Squares-Discriminate Analysis (PLS-DA) were used as unsupervised and supervised techniques for the evaluation of rainbow trout freshness. Results obtained by PCA technique indicated that four classes of samples can be detected using the Vis–NIR mean spectrum by applying a second derivative (D2) preprocessing method. The RCV2 and RPre with D2 preprocessing were 0.97 and 0.98 for Vis–NIR and 0.84 and 0.67 for SWIR, respectively. The corresponding values of RMSECV and RMSEPre were 0.16 and 0.14 in Vis–NIR and 0.44 and 0.76 in SWIR, respectively. Classification model achieved an overall correct classification of 100% and 75% for Vis–NIR and SWIR, respectively. The obtained results using both PCA and PLS-DA methods indicated that the Vis–NIR imaging system performs better than SWIR. Among all applied preprocessing techniques, the second derivative preprocessing achieved the best performance.