Non-destructive determination of water-holding capacity in fresh beef by using NIR hyperspectral imaging

Prediction of freezing point and moisture distribution of beef with dual freeze-thaw cycles using hyperspectral imaging

The freezing point (FP) is an important quality indicator of the superchilled meat. Currently, the potential of hyperspectral imaging (HSI) for predicting beef FP as affected by multiple freeze-thaw (F-T) cycles was explored. Correlation analysis revealed that the FP had a negative correlation with the proportion of bound water (P21) and a positive correlation with the proportion of immobilized water (P22). Moreover, the optimal wavelengths were selected by principal component analysis (PCA). Principal component regression (PCR) and partial least squares regression (PLSR) models were successfully developed based on the optimal wavelengths for predicting FP with determination coefficient in prediction (RP2) of 0.76, 0.76 and root mean square errors in prediction (RMSEP) of 0.12, 0.12, respectively. Additionally, PLSR based on full wavelengths was established for predicting P21 with RP2 of 0.80 and RMSEP of 0.67, and PLSR based on the optimal wavelengths was established for predicting P22 with RP2 of 0.87 and RMSEP of 0.66. The results show the potential of hyperspectral technology to predict the FP and moisture distribution of meat as a nondestructive method.

Water properties in intact wooden breast fillets during refrigerated storage

The wooden breast (WB) condition notably alters moisture content and water holding capacity (WHC) in broiler breast fillets. The purpose of this study was to investigate water properties during refrigerated storage from 4 h to 168 h postmortem using time domain nuclear magnetic resonance (TD-NMR). Water properties measured included mobility (T), proportion (P), and abundance per 100 g of meat (A). Changes in meat quality indicators including compression force, color, pH, cumulative purge loss, and proximate composition were also measured. Compression force and energy of the WB fillets were higher than normal fillets (P < 0.05). Slopes of changes in lightness of the WB and normal fillets were different in skin and bone side (P < 0.05). The slope of the purge loss from the WB fillets was higher than the normal fillets (P < 0.05). Time domain nuclear magnetic resonance analysis showed 4 water populations in intact broiler fillets with transverse relaxation time (T2) constants at approximately 4 to 5 milliseconds (ms) (designated as 2b, corresponding to hydration water or bound water), 40 to 60 ms (designated as 21, corresponding to intra-myofibrillar water or immobilized water), 80 to 210 ms (designated as 22a, corresponding to extra-myofibrillar water or free water with lower mobility) and 210 to 500 ms (designated as 22b, corresponding to extra-myofibrillar water or free water with higher mobility) during early postmortem storage (between 4 h and 72 h postmortem) and only 3 populations (2b, 21, and 22a) after 72 h postmortem. There were interaction effects (P < 0.05) between storage time and WB condition for all water properties except T2b, A2b/100 g, and T22b. The linear change of T21, P21, A21/100 g, T22a, A22a/100 g, P22b, and A22b/100 g in stored WB samples were different from the normal fillets (P < 0.05). During storage, P21 and A21/100 g of the WB fillets exhibited faster linear increases than those of the normal fillets, whereas T21 and T22a of the normal fillets and A22a/100 g, P22b, and A22b/100 g of the WB fillets showed faster linear decreases (P < 0.05). Our data demonstrate that the WB condition affects changes in water properties in broiler fillets during postmortem refrigerated storage.

Predictive Analysis of Linoleic Acid in Red Meat Employing Advanced Ensemble Models of Bayesian and CNN-Bi-LSTM Decision Layer Fusion Based Hyperspectral Imaging

Rapid non-destructive testing technologies are effectively used to analyze and evaluate the linoleic acid content while processing fresh meat products. In current study, hyperspectral imaging (HSI) technology was combined with deep learning optimization algorithm to model and analyze the linoleic acid content in 252 mixed red meat samples. A comparative study was conducted by experimenting mixed sample data preprocessing methods and feature wavelength extraction methods depending on the distribution of linoleic acid content. Initially, convolutional neural network Bi-directional long short-term memory (CNN-Bi-LSTM) model was constructed to reduce the loss of the fully connected layer extracted feature information and optimize the prediction effect. In addition, the prediction process of overfitting phenomenon in the CNN-Bi-LSTM model was also targeted. The Bayesian-CNN-Bi-LSTM (Bayes-CNN-Bi-LSTM) model was proposed to improve the linoleic acid prediction in red meat through iterative optimization of Gaussian process acceleration function. Results showed that best preprocessing effect was achieved by using the detrending algorithm, while 11 feature wavelengths extracted by variable combination population analysis (VCPA) method effectively contained characteristic group information of linoleic acid. The Bi-directional LSTM (Bi-LSTM) model combined with the feature extraction data set of VCPA method predicted 0.860 Rp2 value of linoleic acid content in red meat. The CNN-Bi-LSTM model achieved an Rp2 of 0.889, and the optimized Bayes-CNN-Bi-LSTM model was constructed to achieve the best prediction with an Rp2 of 0.909. This study provided a reference for the rapid synchronous detection of mixed sample indicators, and a theoretical basis for the development of hyperspectral on-line detection equipment.

Non-destructive detection of protein content in mulberry leaves by using hyperspectral imaging

Protein content is one of the most important indicators for assessing the quality of mulberry leaves. This work is carried out for the rapid and non-destructive detection of protein content of mulberry leaves using hyperspectral imaging (HSI) (Specim FX10 and FX17, Spectral Imaging Ltd., Oulu, Finland). The spectral range of the HSI acquisition system and data processing methods (pretreatment, feature extraction, and modeling) is compared. Hyperspectral images of three spectral ranges in 400-1,000 nm (Spectral Range I), 900-1,700 nm (Spectral Range II), and 400-1,700 nm (Spectral Range III) were considered. With standard normal variate (SNV), Savitzky-Golay first-order derivation, and multiplicative scatter correction used to preprocess the spectral data, and successive projections algorithm (SPA), competitive adaptive reweighted sampling, and random frog used to extract the characteristic wavelengths, regression models are constructed by using partial least square and least squares-support vector machine (LS-SVM). The protein content distribution of mulberry leaves is visualized based on the best model. The results show that the best results are obtained with the application of the model constructed by combining SNV with SPA and LS-SVM, showing an R 2 of up to 0.93, an RMSE of just 0.71 g/100 g, and an RPD of up to 3.83 based on the HSI acquisition system of 900-1700 nm. The protein content distribution map of mulberry leaves shows that the protein of healthy mulberry leaves distributes evenly among the mesophyll, with less protein content in the vein of the leaves. The above results show that rapid, non-destructive, and high-precision detection of protein content of mulberry leaves can be achieved by applying the SWIR HSI acquisition system combined with the SNV-SPA-LS-SVM algorithm.

Deep Learning-Based Near-Infrared Hyperspectral Imaging for Food Nutrition Estimation

The limited nutritional information provided by external food representations has constrained the further development of food nutrition estimation. Near-infrared hyperspectral imaging (NIR-HSI) technology can capture food chemical characteristics directly related to nutrition and is widely used in food science. However, conventional data analysis methods may lack the capability of modeling complex nonlinear relations between spectral information and nutrition content. Therefore, we initiated this study to explore the feasibility of integrating deep learning with NIR-HSI for food nutrition estimation. Inspired by reinforcement learning, we proposed OptmWave, an approach that can perform modeling and wavelength selection simultaneously. It achieved the highest accuracy on our constructed scrambled eggs with tomatoes dataset, with a determination coefficient of 0.9913 and a root mean square error (RMSE) of 0.3548. The interpretability of our selection results was confirmed through spectral analysis, validating the feasibility of deep learning-based NIR-HSI in food nutrition estimation.

Evaluation of Dual-Band Near-Infrared Spectroscopy and Chemometric Analysis for Rapid Quantification of Multi-Quality Parameters of Soy Sauce Stewed Meat

The objective of this study was to evaluate the performance of near-infrared spectroscopy (NIRS) systems operated in dual band for the non-destructive measurement of the fat, protein, collagen, ash, and Na contents of soy sauce stewed meat (SSSM). Spectra in the waveband ranges of 650-950 nm and 960-1660 nm were acquired from vacuum-packed ready-to-eat samples that were purchased from 97 different brands. Partial least squares regression (PLSR) was employed to develop models predicting the five critical quality parameters. The results showed the best predictions were for the fat (Rp = 0.808; RMSEP = 2.013 g/kg; RPD = 1.666) and protein (Rp = 0.863; RMSEP = 3.372 g/kg; RPD = 1.863) contents, while barely sufficient performances were found for the collagen (Rp = 0.524; RMSEP = 1.970 g/kg; RPD = 0.936), ash (Rp = 0.384; RMSEP = 0.524 g/kg; RPD = 0.953), and Na (Rp = 0.242; RMSEP = 2.097 g/kg; RPD = 1.042) contents of the SSSM. The quality of the content predicted by the spectrum of 960-1660 nm was generally better than that for the 650-950 nm range, which was retained in the further prediction of fat and protein. To simplify the models and make them practical, regression models were established using a few wavelengths selected by the random frog (RF) or regression coefficients (RCs) method. Consequently, ten wavelengths (1048 nm, 1051 nm, 1184 nm, 1191 nm, 1222 nm, 1225 nm, 1228 nm, 1450 nm, 1456 nm, 1510 nm) selected by RF and eight wavelengths (1019 nm, 1097 nm, 1160 nm, 1194 nm, 1245 nm, 1413 nm, 1441 nm, 1489 nm) selected by RCs were individually chosen for the fat and protein contents to build multi-spectral PLSR models. New models led to the best predictive ability of Rp, RMSEP, and RPD of 0.812 and 0.855, 1.930 g/kg and 3.367 g/kg, and 1.737 and 1.866, respectively. These two simplified models both yielded comparable performances to their corresponding full-spectra models, demonstrating the effectiveness of these selected variables. The overall results indicate that NIRS, especially in the spectral range of 960-1660 nm, is a potential tool in the rapid estimation of the fat and protein contents of SSSM, while not providing particularly good prediction statistics for collagen, ash, and Na contents.

Artificial Intelligence in Food Safety: A Decade Review and Bibliometric Analysis

Artificial Intelligence (AI) technologies have been powerful solutions used to improve food yield, quality, and nutrition, increase safety and traceability while decreasing resource consumption, and eliminate food waste. Compared with several qualitative reviews on AI in food safety, we conducted an in-depth quantitative and systematic review based on the Core Collection database of WoS (Web of Science). To discover the historical trajectory and identify future trends, we analysed the literature concerning AI technologies in food safety from 2012 to 2022 by CiteSpace. In this review, we used bibliometric methods to describe the development of AI in food safety, including performance analysis, science mapping, and network analysis by CiteSpace. Among the 1855 selected articles, China and the United States contributed the most literature, and the Chinese Academy of Sciences released the largest number of relevant articles. Among all the journals in this field, PLoS ONE and Computers and Electronics in Agriculture ranked first and second in terms of annual publications and co-citation frequency. The present character, hot spots, and future research trends of AI technologies in food safety research were determined. Furthermore, based on our analyses, we provide researchers, practitioners, and policymakers with the big picture of research on AI in food safety across the whole process, from precision agriculture to precision nutrition, through 28 enlightening articles.

Real-time monitoring the color changes of large yellow croaker (Larimichthys crocea) fillets based on hyperspectral imaging empowered with artificial intelligence

Vis-NIR hyperspectral imaging (HSI) system combined with artificial neural networks was investigated for the first time to monitor color changes of large yellow croaker (Larimichthys crocea) fillets during low-temperature storage. Feed-forward neural networks (FNN) empowered with the leaky rectified linear unit (Leaky-Relu) have been developed as a non-linear quantitative analysis model. It presented accurate predictive power for color changes based on optimal spectra (with R²_P of 0.908, 0.915, and 0.977; and RMSEP of 1.062, 3.315, and 0.082 for L*, a*, and b*, respectively). In final, the simplified FNN-Leaky-Relu model (S-FNN-L) was utilized to visualize the distribution maps of color parameters in the fillets. The results demonstrated the feasibility of HSI could replace the traditional colorimeter to determine the spatial distribution in the color measurement of fish fillets with a rapid and non-invasive technique.

UV-fluorescence imaging for real-time non-destructive monitoring of pork freshness

This study aimed to develop a cost-effective fluorescence imaging system to rapidly monitor pork freshness indicators during chilled storage. The system acquired fluorescence images of pork and the color features were extracted from these images to establish partial least squares regression (PLSR) models to predict total volatile basic nitrogen (TVB-N), total viable count (TVC), pH for pork. For TVB-N, TVC and pH values, R_p were 0.92, 0.88 and 0.74, residual predictive deviation (RPD) were 2.24, 2.03, and 1.19, respectively. For TVB-N and TVC indicators showed that the predictive ability of this model was largely comparable to that of fluorescence hyperspectral imaging. However, combining fluorescence and color imaging improved the model's predictive ability. For TVB-N, TVC and pH, R_p were 0.94, 0.93 and 0.85, RPD were 2.62, 2.59, and 1.95, respectively. Therefore, this study developed a system with great potential for detecting the value of most pork quality indicators in real-time.

The application of computer vision systems in meat science and industry - A review

Computer vision systems (CVS) are applied to macro- and microscopic digital photographs captured using digital cameras, ultrasound scanners, computer tomography, and wide-angle imaging cameras. Diverse image acquisition devices make it technically feasible to obtain information about both the external features and internal structures of targeted objects. Attributes measured in CVS can be used to evaluate meat quality. CVS are also used in research related to assessing the composition of animal carcasses, which might help determine the impact of cross-breeding or rearing systems on the quality of meat. The results obtained by the CVS technique also contribute to assessing the impact of technological treatments on the quality of raw and cooked meat. CVS have many positive attributes including objectivity, non-invasiveness, speed, and low cost of analysis and systems are under constant development an improvement. The present review covers computer vision system techniques, stages of measurements, and possibilities for using these to assess carcass and meat quality.

Blue Light for Inactivation of Meatborne Pathogens and Maintaining the Freshness of Beef

ABSTRACT

Beef is rich in various nutrients but easily spoils due to bacterial contamination; thus, a bactericidal method is needed to inactivate meatborne pathogens while maintaining the freshness of beef. The present study was conducted to investigate for the first time the bactericidal effect of blue light (BL) at 415 nm against four meatborne pathogens (methicillin-resistant Staphylococcus aureus, Escherichia coli, Salmonella Typhimurium, and Listeria monocytogenes) both in vitro and inoculated onto the surface of fresh beef. The populations of the four pathogens on the nonirradiated control beef did not change significantly (P > 0.05), whereas a dose-dependent inactivation effect was found for BL-treated beef both in vitro and in vivo. On the beef cuts, BL at 109.44 J/cm2 inactivated 90% of inoculated cells of the tested strains (P < 0.05), and this inactivation effect was sustained during 7 days of cold storage. Insignificant changes in lipid oxidation rate, water holding capacity, and cooking loss were found during storage between the control beef and the beef irradiated at 109.44 J/cm2 at the same time. BL had a minor and nonsignificant effect on surface color and free amino acid concentrations. The pH of the treated beef increased more slowly (P < 0.05) than did that of untreated beef. These results suggest that BL could be a novel bactericide and could help maintain the freshness of beef.

HIGHLIGHTS

Detection of Foreign Materials on Broiler Breast Meat Using a Fusion of Visible Near-Infrared and Short-Wave Infrared Hyperspectral Imaging

Foreign material (FM) found on a poultry product lowers the quality and safety of the product. We developed a fusion method combining two hyperspectral imaging (HSI) modalities in the visible-near infrared (VNIR) range of 400–1000 nm and the short-wave infrared (SWIR) range of 1000–2500 nm for the detection of FMs on the surface of fresh raw broiler breast fillets. Thirty different types of FMs that could be commonly found in poultry processing plants were used as samples and prepared in two different sizes (5 × 5 mm2 and 2 × 2 mm2). The accuracies of the developed Fusion model for detecting 2 × 2 mm2 pieces of polymer, wood, and metal were 95%, 95%, and 81%, respectively, while the detection accuracies of the Fusion model for detecting 5 × 5 mm2 pieces of polymer, wood, and metal were all 100%. The performance of the Fusion model was higher than the VNIR- and SWIR-based detection models by 18% and 5%, respectively, when F1 scores were compared, and by 38% and 5%, when average detection rates were compared. The study results suggested that the fusion of two HSI modalities could detect FMs more effectively than a single HSI modality.

Handheld near-infrared spectrometer allows on-line prediction of beef quality traits

The aim of this study was to evaluate the ability of a miniaturized near-infrared spectrometer to predict chemical parameters, technological and quality traits, fatty acids and minerals in intact Longissimus thoracis and Trapezius obtained from the ribs of 40 Charolais cattle. Modified partial least squares regression analysis to correlate spectra information to reference values, and several scatter correction and mathematical treatments have been tested. Leave-one-out cross-validation results showed that the handheld instrument could be used to obtain a good prediction of moisture and an approximate quantitative prediction of fat or protein contents, a*, b*, shear force and purge loss with coefficients of determination above 0.66. Moreover, prediction models were satisfactory for proportions of MUFA, PUFA, oleic and palmitic acids, for Fe and Cu contents. Overall, results exhibited the usefulness of the on-line miniaturized tool to predict some beef quality traits and the possibility to use it with commercial cuts without sampling, carcass deterioration nor grinding and consequent meat products' loss.

Modeling relationship between protein oxidation and denaturation and texture, moisture loss of bighead carp (Aristichthys Nobilis) during frozen storage

To evaluate the effects of protein oxidation and denaturation on the fish texture and moisture loss during frozen storage, we measured the changes of protein oxidation and denaturation (salt-soluble protein (SSP), total sulfhydryl (SH), disulfide (SS), carbonyl contents and Ca²⁺ -ATPase activity), texture (hardness), and moisture loss (drip loss) of bighead carp fillets stored at -12, -20 and -28°C during 16 weeks. These data were employed to develop partial least squares regression (PLSR) model, radial basis function neural network (RBFNN) model, PLSR-RBFNN (PR) model and RBFNN-PLSR (RP) model. The results showed that the RP model provided no enhancement to RBFNN model because it had the exactly same root mean square error (RMSE) and R² . PLSR model showed better performance than other models when predicting hardness. More appropriate linear or linearity-dominant hybrid model needed to be explored to establish the relationship between protein oxidation and denaturation and texture. The PR model performed better than other models in predicting drip loss with its lower RMSE and higher R² , which revealed both linear and nonlinear relationship between protein oxidation and denaturation and moisture loss. Therefore, the PR model was a promising and encouraging tool to provide a more comprehensive understanding of the relationship between protein oxidation and denaturation and moisture loss of fish during frozen storage. PRACTICAL APPLICATION: The study explored the effects of protein oxidation and denaturation on the texture and moisture loss of bighead carp during frozen storage (-12 to -28°C). PLSR model showed better performance than other models when predicting the relationship between protein oxidation and denaturation and texture. The PR model was an available tool for manufacturers to predict the relationship between protein oxidation and denaturation and moisture loss.

Chemometrics in Tandem with Hyperspectral Imaging for Detecting Authentication of Raw and Cooked Mutton Rolls

Authentication assurance of meat or meat products is critical in the meat industry. Various methods including DNA- or protein-based techniques are accurate for assessing meat authenticity, however, they are destructive, expensive, or laborious. This study explores the feasibility of chemometrics in tandem with hyperspectral imaging (HSI) for identifying raw and cooked mutton rolls substitution by pork and duck rolls. Raw or cooked samples (n = 180) of three meat species were prepared to collect hyperspectral images in range of 400-1000 nm. Spectra were extracted from representative regions of interest (ROIs), and spectral principal component analysis (PCA) revealed that PC1 and PC2 were effective for the identification. Different methods including standard normal variable (SNV), first and second derivatives, and normalization were individually employed for spectral preprocessing, and modeling methods of partial least squares-discriminant analysis (PLS-DA) and support vector machines (SVM) were also individually applied to develop classification models for both the raw and the cooked. Results showed that PLS-DA model developed by raw spectra presented the highest 100% correct classification rate (CCR) of success in all sets. After that, effective wavelengths selected by successive projections algorithm (SPA) built optimal simplified models which didn't influence the modeling results compared with full spectra regardless of the meat roll states. Therefore, SPA-PLS-DA models were subsequently used to visualize the raw and cooked meat rolls classification. As a consequence, the general meat species of both raw and cooked meat rolls were readily discernible in pixel-wise manner by generating classification maps. The results showed that HSI combined with chemometrics can be used to identify the authentication of raw and cooked mutton rolls substituted by pork and duck rolls accurately. This promising methodology provides a reference which can be extended to the classification or grading of other meat rolls.

Application of principal component analysis with instrumental analysis and sensory evaluation for assessment of chicken breast meat juiciness

1. The objectives of this study were to use principal component analysis (PCA) to analyse the variability of the three instrumental and 14 descriptive sensory properties of chicken breast meat. The meat was cooked until the internal temperature reached 85°C and further cooked for 0, 20, and 40 min. The second objective was to identify the most critical variables for assessing meat juiciness.2. Cooking loss and moisture content exhibited high correlation with sensorial moisture release and mouth feel.3. The distribution of objects on the axes of the first two principal components (PCs) enabled the identification of three groups undergoing different cooking durations. The four major PCs explained 80.0% of the total variability.4. Cooking loss, moisture content, water-holding capacity, sensorial moisture release and mouth feel were demonstrated as the most effective variables for the first two PCs. PCA with instrumental and sensory analyses proved an effective procedure for systematically and comprehensively judging chicken meat juiciness.

A Review on Meat Quality Evaluation Methods Based on Non-Destructive Computer Vision and Artificial Intelligence Technologies

Increasing meat demand in terms of both quality and quantity in conjunction with feeding a growing population has resulted in regulatory agencies imposing stringent guidelines on meat quality and safety. Objective and accurate rapid non-destructive detection methods and evaluation techniques based on artificial intelligence have become the research hotspot in recent years and have been widely applied in the meat industry. Therefore, this review surveyed the key technologies of non-destructive detection for meat quality, mainly including ultrasonic technology, machine (computer) vision technology, near-infrared spectroscopy technology, hyperspectral technology, Raman spectra technology, and electronic nose/tongue. The technical characteristics and evaluation methods were compared and analyzed; the practical applications of non-destructive detection technologies in meat quality assessment were explored; and the current challenges and future research directions were discussed. The literature presented in this review clearly demonstrate that previous research on non-destructive technologies are of great significance to ensure consumers’ urgent demand for high-quality meat by promoting automatic, real-time inspection and quality control in meat production. In the near future, with ever-growing application requirements and research developments, it is a trend to integrate such systems to provide effective solutions for various grain quality evaluation applications.

Variation in the terminology and methodologies applied to the analysis of water holding capacity in meat research

Studies examining meat quality variation, possibly resulting from animal physiology, processing, or ingredient additions, are likely to include at least one measure of water holding capacity (WHC). Methods for evaluating WHC can be classified as direct or indirect. Direct methods either gauge natural release of fluids from muscle or require the application of force to express water. The indirect methods do not actually measure WHC. They attempt to separate meat into two or three categories based on predictions of direct method results: the extreme of high and low WHC and an optional 'normal' group. Considerable statistical analyses are required to generate these predictive models. Presently, there are inconsistent terms (e.g., water holding, WHC, water binding, water binding potential/capacity) used to describe WHC and no standardized techniques recommended to evaluate it. To ensure that results can be compared across different laboratories, a better consensus must be reached in how these terms are employed and how this critical parameter is determined.

Near-infrared hyperspectral imaging for detection and visualization of offal adulteration in ground pork

Hyperspectral imaging (HSI) technique was investigated to explore a feasible protocol for detecting the potential offal (lung) adulteration in ground pork. Tested samples (176 adulterated and 2 controls) were first prepared with adulterant of ground lung in range of 0%-100% (w/w) at 10% intervals. After hyperspectral images were acquired and calibrated in reflectance mode (400-1000 nm), full spectra were extracted from identified regions of interests (ROIs) and then transformed into absorbance and Kubelka-Munck spectral units, respectively. Partial least squares regression (PLSR) models based on full spectra showed that raw reflectance spectra with no preprocessings performed best with coefficient of determination (R_p²) of 0.98, root mean square error (RMSEP) of 4.25%, and ratio performance deviation (RPD) of 7.53 in prediction set. To reduce the high dimensionality of spectra, data was further explored using principal component loadings, two-dimensional correlation spectroscopy (2D-COS), and regression coefficients (RC), respectively. The optimal performance of established simplified PLSR model were acquired using eleven featured wavelengths selected by PC loadings with R_p² of 0.98, RMSEP of 4.47% and RPD of 7.16. Finally, the limit of detection (LOD) was calculated to be a satisfactory 7.58%, and readily discernible visualization procedure using preferred simplified PLSR model yielded satisfactory spatial distribution of adulteration situation. Control samples with known distribution were also visualized to successfully prove the validity. Consequently, this research offers an alternative assay for visually and rapidly detecting offal of lung adulteration in ground pork.

Heavy metal Hg stress detection in tobacco plant using hyperspectral sensing and data-driven machine learning methods

Accurate detection of heavy metal stress on the growth status of plants is of great concern for agricultural production and management, food security, and ecological environment. A proximal hyperspectral imaging (HSI) system covered the visible/near-infrared (Vis/NIR) region of 400-1000 nm coupled with machine learning methods were employed to discriminate the tobacco plants stressed by different concentration of heavy metal Hg. After acquiring hyperspectral images of tobacco plants stressed by heavy metal Hg with concentration solutions of 0 mg·L-1 (non-stressed groups), 1, 3, and 5 mg·L-1 (3 stressed groups), regions of interest (ROIs) of canopy in tobacco plants were identified for spectra processing. Meanwhile, tobacco plant's appearance and microstructure of mesophyll tissue in tobacco leaves were analyzed. After that, clustering effects of the non-stressed and stressed groups were revealed by score plots and score images calculated by principal component analysis (PCA). Then, loadings of PCA and competitive adaptive reweighted sampling (CARS) algorithm were employed to pick effective wavelengths (EWs) for discriminating non-stressed and stressed samples. Partial least squares discriminant analysis (PLS-DA) and least-squares support vector machine (LS-SVM) were utilized to estimate the stressed tobacco plants status with different concentrations Hg solutions. The performances of those models were evaluated using confusion matrixes (CMes) and receiver operating characteristics (ROC) curves. Results demonstrated that PLS-DA models failed to offer relatively good result, and this algorithm was abandoned to classify the stressed and non-stressed groups of tobacco plants. Compared to LS-SVM model based on full spectra (FS-LS-SVM), the LS-SVM model established EWs selected by CARS (CARS-LS-SVM) carried 13 variables provided an accuracy of 100%, which was promising to achieve the qualitative discrimination of the non-stressed and stressed tobacco plants. Meanwhile, for revealing the discrepancy between 3 stressed groups of tobacco plants, the other FS-LS-SVM, PCA-LS-SVM, and CARS-LS-SVM models were setup and offered relatively low accuracies of 55.56%, 51.11% and 66.67%, respectively. Performance of those 3 LS-SVM discriminative models was also poorly performing to differentiate 3 stressed groups of tobacco plants, which might be caused by low concentration of heavy metal and similar canopy (especially in fresh leaves) of plant. The achievements of the research indicated that HSI coupled with machine learning methods had a powerful potential to discriminate tobacco plant stressed by heavy metal Hg.

A review of the medical hyperspectral imaging systems and unmixing algorithms' in biological tissues

Hyperspectral fluorescence imaging (HFI) is a well-known technique in the medical research field and is considered a non-invasive tool for tissue diagnosis. This review article gives a brief introduction to acquisition methods, including the image preprocessing methods, feature selection and extraction methods, data classification techniques and medical image analysis along with recent relevant references. The process of fusion of unsupervised unmixing techniques with other classification methods, like the combination of support vector machine with an artificial neural network, the latest snapshot Hyperspectral imaging (HSI) and vortex analysis techniques are also outlined. Finally, the recent applications of hyperspectral images in cellular differentiation of various types of cancer are discussed.

Quality Assessment of Beef Using Computer Vision Technology

Imaging technique or computer vision (CV) technology has received huge attention as a rapid and non-destructive technique throughout the world for measuring quality attributes of agricultural products including meat and meat products. This study was conducted to test the ability of CV technology to predict the quality attributes of beef. Images were captured from longissimus dorsi muscle in beef at 24 h post-mortem. Traits evaluated were color value (L*, a*, b*), pH, drip loss, cooking loss, dry matter, moisture, crude protein, fat, ash, thiobarbituric acid reactive substance (TBARS), peroxide value (POV), free fatty acid (FFA), total coliform count (TCC), total viable count (TVC) and total yeast-mould count (TYMC). Images were analyzed using the Matlab software (R2015a). Different reference values were determined by physicochemical, proximate, biochemical and microbiological test. All determination were done in triplicate and the mean value was reported. Data analysis was carried out using the programme Statgraphics Centurion XVI. Calibration and validation model were fitted using the software Unscrambler X version 9.7. A higher correlation found in a* (r=0.65) and moisture (r=0.56) with 'a*' value obtained from image analysis and the highest calibration and prediction accuracy was found in lightness (r² _c=0.73, r² _p=0.69) in beef. Results of this work show that CV technology may be a useful tool for predicting meat quality traits in the laboratory and meat processing industries.

Cultivar Discrimination of Single Alfalfa (Medicago sativa L.) Seed via Multispectral Imaging Combined with Multivariate Analysis

Rapid and accurate discrimination of alfalfa cultivars is crucial for producers, consumers, and market regulators. However, the conventional routine of alfalfa cultivars discrimination is time-consuming and labor-intensive. In this study, the potential of a new method was evaluated that used multispectral imaging combined with object-wise multivariate image analysis to distinguish alfalfa cultivars with a single seed. Three multivariate analysis methods including principal component analysis (PCA), linear discrimination analysis (LDA), and support vector machines (SVM) were applied to distinguish seeds of 12 alfalfa cultivars based on their morphological and spectral traits. The results showed that the combination of morphological features and spectral data could provide an exceedingly concise process to classify alfalfa seeds of different cultivars with multivariate analysis, while it failed to make the classification with only seed morphological features. Seed classification accuracy of the testing sets was 91.53% for LDA, and 93.47% for SVM. Thus, multispectral imaging combined with multivariate analysis could provide a simple, robust and nondestructive method to distinguish alfalfa seed cultivars.

Application of hyperspectral technology in detection of agricultural products and food: A Review

Food is the foundation of human survival. With the development and progress of society, people increasingly focus on the problems of food quality and safety, which is closely related to human's health. Thus, the whole industrial chain from farmland to dining table need to be strictly controlled. Traditional detection methods are time-consuming, laborious, and destructive. In recent years, hyperspectral technology has been more and more applied to food safety and quality detection, because the technology can achieve rapid and nondestructive detection of food, and the requirement to experimental condition is low; operability is strong. In this paper, hyperspectral imaging technology was briefly introduced, and its application in agricultural products and food detection in recent years was systematically summarized, and the key points in the research process were deeply discussed. This work lays a solid foundation for the peers to the following in-depth research and application of this technology.

Chemometric strategies for nondestructive and rapid assessment of nitrate content in harvested spinach using Vis-NIR spectroscopy

The overuse of nitrogenous fertilizers leads to an increase in the nitrate content of green leafy vegetables. Consumption of food with excess nitrate is not advisable because it results in human ailment. In this study, spinach leaves were harvested from plants grown under nine varying (0 to 400 kg/ha) nitrogenous fertilizer doses. A total of 261 samples were used to predict the nitrate content in spinach leaves using Vis-NIR (350 to 2,500 nm). The nitrate content was measured destructively using the ion-selective conductive method. Partial least square (PLS) regression models were developed using whole spectra and featured wavelengths. Spectral data were pre-processed using different spectral pre-processing techniques such as Savitzky-Golay (SG) derivative, standard normal variate (SNV), multiplicative scatter correction (MSC), baseline correction, and detrending. The predictive accuracy of the PLS model had improved after pre-processing of spectral data with MSC (RPD_CV = 1.767; SE_CV = 545.745; bias_CV = -3.107; slope_CV = 0.698) and SNV (RPD_CV = 1.768; SE_CV = 545.337; bias_CV = -3.201; slope_CV = 0.698) technique, but this was not significant (P < 0.05) as compared with raw spectral data (RPD_CV = 1.679; SE_CV = 572.669; bias_CV = -7.046; slope_CV = 0.687). The effective wavelengths for measurement nitrate content in spinach leaves were identified as 558, 706, 780, 1,000, and 1,420 nm. The performance of PLS model developed with effective wavelengths also had good prediction accuracy (RPD_CV = 1.482; SE_CV = 648.672; bias_CV = -3.805; slope_CV = 0.565) but significantly lower than the performance of model developed with full spectral data. The overall results of this study suggest that Vis-NIR spectroscopy can be an important tool and has great potential for the rapid and nondestructive assessment of nitrate content in harvested spinach, with a view to ascertain the suitability of the harvest for food uses. PRACTICAL APPLICATION: Better production and brighter color of leafy vegetable drive the farming community to overuse nitrogenous fertilizer. This has resulted in higher nitrate content in vegetables. It has been widely reported that consumption of these vegetables has carcinogenic effects on human beings. The prediction of nitrate content in leafy vegetables by traditional methods is time-consuming (30 min, including sample preparation time), destructive, and tedious; moreover, it cannot be used for inline applications. This study reports spectroscopy-based rapid (<5 s) assessment technique for nitrate measurement. A multivariable PLS model was developed using wavelengths representing nitrate content. This model can be adopted by food industries for inline applications.

Rapid detection of adulteration of minced beef using Vis/NIR reflectance spectroscopy with multivariate methods

High economic returns induce the continuous occurrence of meat adulteration. In this study, visible/near-infrared (Vis/NIR) reflectance spectroscopy with multivariate methods was used for the rapid detection of adulteration in minced beef. First, the reflectance spectra of different adulterated minced beef samples were measured at 350-2500 nm. Standardization and Savitzky-Golay (SG) smoothing were applied to reduce spectral interference and noise. Then, support vector machine (SVM), random forest (RF), partial least squares regression (PLSR), and deep convolutional neural network (DCNN) were adopted for adulteration type identification and level prediction. Moreover, principal component analysis (PCA), locally linear embedding (LLE), subwindow permutation analysis (SPA), and competitive adaptive reweighted sampling (CARS) were performed to eliminate redundant information. SG smoothing performed better on interference reduction. DCNN and PCA identified adulteration type with the accuracy above 99%. In adulteration level prediction, the RF with spectra of important wavelengths selected by CARS provided optimal performance for beef adulterated with pork, and coefficient of determination of prediction (R²_P) and the root mean square error of prediction (RMSEP) were 0.973 and 2.145. The best prediction for beef adulterated with beef heart was obtained using PLSR and CARS with R²_P of 0.960 and RMSEP of 2.758. Accordingly, Vis/NIR reflectance spectroscopy coupled with multivariate methods can provide the rapid and accurate detection of adulterated minced beef.

Non-destructive identification of single hard seed via multispectral imaging analysis in six legume species

BACKGROUND

Physical dormancy (hard seed) occurs in most species of Leguminosae family and has great consequences not only for ecological adaptation but also for agricultural practice of these species. A rapid, nondestructive and on-site screening method to detect hard seed within species is fundamental important for maintaining seed vigor and germplasm storage as well as understanding seed adaptation to various environment. In this study, the potential of multispectral imaging with object-wise multivariate image analysis was evaluated as a way to identify hard and soft seeds in Acacia seyal, Galega orientulis, Glycyrrhiza glabra, Medicago sativa, Melilotus officinalis, and Thermopsis lanceolata. Principal component analysis (PCA), linear discrimination analysis (LDA), and support vector machines (SVM) methods were applied to classify hard and soft seeds according to their morphological features and spectral traits.

RESULTS

The performance of discrimination model via multispectral imaging analysis was varied with species. For M. officinalis, M. sativa, and G. orientulis, an excellent classification could be achieved in an independent validation data set. LDA model had the best calibration and validation abilities with the accuracy up to 90% for M. sativa. SVM got excellent seed discrimination results with classification accuracy of 91.67% and 87.5% for M. officinalis and G. orientulis, respectively. However, both LDA and SVM model failed to discriminate hard and soft seeds in A. seyal, G. glabra, and T. lanceolate.

CONCLUSIONS

Multispectral imaging together with multivariate analysis could be a promising technique to identify single hard seed in some legume species with high efficiency. More legume species with physical dormancy need to be studied in future research to extend the use of multispectral imaging techniques.

Meat quality evaluation based on computer vision technique: A review

Nowadays people tend to include more meat in their diet thanks to the improvement in standards of living as well as an increase in awareness of meat nutritive values. To ensure public health, therefore, there is a need for a rise in worldwide meat production and consumption. Further attention is also required as to how the safety and the quality of meat production process should be assessed. Classical methods of meat quality assessment, however, have some disadvantages; expensive and time-consuming. This study intends to introduce an alternative method known as Computer Vision (CV) for the assessment of various quality parameters of muscle foods. CV has several advantages over the traditional methods. It is non-destructive, easy, and quick, hence, more efficient in meat quality assessments. This study aims to investigate different quality characteristics of some muscle foods using CV. It closes with a discussion on the future challenges and expected opportunities of the practical application of CV in the meat industry.

Integration of spectral and textural features of visible and near-infrared hyperspectral imaging for differentiating between normal and white striping broiler breast meat

White striping (WS), an emerging muscle myopathy in poultry industry, is gaining increasing attention globally. In this study, visible and near-infrared hyperspectral imaging (HSI, 400-1000 nm) was investigated for developing an optical sensing technique to differentiate WS broiler breast fillets (pectoralis major) from normal fillets. The minimum noise fraction (MNF), followed by an inverse MNF (IMNF), was conducted to improve the signal-to-noise ratio of hyperspectral images during the pre-processing process. Three regions of interest (ROIs) were selected at cranial, middle and caudal locations within each fillet image. Spectral principal component analysis (PCA) revealed that PC₂ and PC₃ were effective for the differentiation and key wavelengths (450, 492, 541, 581, 629, 869 and 980 nm) were selected from the corresponding PC loadings. Spatial texture features on corresponding score images were obtained using gray level co-occurrence matrix (GLCM) and grayscale histogram statistics (GHS), respectively. Partial least squares discriminant analysis (PLS-DA) models were evaluated with various inputs including spectral (full and key wavelengths), textural and fused features. GLCM features improved performance of multispectral imaging with the highest correct classification rate (CCR) of 91.7%, AUC value (0.917), and Kappa coefficient (0.833) for prediction set. Considering the complexity and heterogeneity of meat samples at different locations, the optimal sampling location was also analyzed and results provided the evidence that the cranial location worked most effectively for the differentiation between normal and WS samples. Overall, results confirmed the great potential of HSI in range of 400-1000 nm in differentiation between normal and WS chicken breast meat.

Development of a Low-Cost Multi-Waveband LED Illumination Imaging Technique for Rapid Evaluation of Fresh Meat Quality

Determining the quality of meat has always been essential for the food industry because consumers prefer superior quality meat. Therefore, the food industry requires the development of a rapid and non-destructive method for meat-quality determination. Over the past few years, a number of techniques have been presented for monitoring meat–chemical attributes. However, most previous techniques are quite expensive, destructive, and require complex hardware to operate. Thus, in this work, we demonstrate a low-cost sensing technique (eliminating the expensive equipment and complicated design) for meat–chemical quality detection. The newly developed system was integrated with a low-cost monochrome camera and ordinary light-emitting diode (LED) light sources, with fifteen different wavebands ranging from 458 to 950 nm. The monochrome camera captures images of the meat sample across a spectral range from 458 to 950 nm using a single snapshot method. The chemical values (e.g., moisture, fat, and protein) were also determined using conventional methods. The collected images were combined to produce a multispectral data cube and to extract spectral data. Partial least squares (PLS) and support vector regression (SVR) modeling were used on the extracted spectra and chemical values. The developed models for meat samples displayed accurate chemical-component prediction ( R 2 > 0.80). Our model, based on a monochrome sensor using only fifteen wavebands, provided reasonable results compared with the previously developed expensive spectroscopic techniques. Therefore, this complementary metal-oxide semiconductor (CMOS) based multispectral sensing technique may have the potential to detect meat quality, thereby facilitating a simple, fast, and cost-effective method applicable to small-scale meat-processing industries.

Utilization of computer vision and multispectral imaging techniques for classification of cowpea (Vigna unguiculata) seeds

BACKGROUND

The traditional methods for evaluating seeds are usually performed through destructive sampling followed by physical, physiological, biochemical and molecular determinations. Whilst proven to be effective, these approaches can be criticized as being destructive, time consuming, labor intensive and requiring experienced seed analysts. Thus, the objective of this study was to investigate the potential of computer vision and multispectral imaging systems supported with multivariate analysis for high-throughput classification of cowpea (Vigna unguiculata) seeds. An automated computer-vision germination system was utilized for uninterrupted monitoring of seeds during imbibition and germination to identify different categories of all individual seeds. By using spectral signatures of single cowpea seeds extracted from multispectral images, different multivariate analysis models based on linear discriminant analysis (LDA) were developed for classifying the seeds into different categories according to ageing, viability, seedling condition and speed of germination.

RESULTS

The results revealed that the LDA models had good accuracy in distinguishing 'Aged' and 'Non-aged' seeds with an overall correct classification (OCC) of 97.51, 96.76 and 97%, 'Germinated' and 'Non-germinated' seeds with OCC of 81.80, 79.05 and 81.0%, 'Early germinated', 'Medium germinated' and 'Dead' seeds with OCC of 77.21, 74.93 and 68.00% and among seeds that give 'Normal' and 'Abnormal' seedlings with OCC of 68.08, 64.34 and 62.00% in training, cross-validation and independent validation data sets, respectively. Image processing routines were also developed to exploit the full power of the multispectral imaging system in visualizing the difference among seed categories by applying the discriminant model in a pixel-wise manner.

CONCLUSION

The results demonstrated the capability of the multispectral imaging system in the ultraviolet, visible and shortwave near infrared range to provide the required information necessary for the discrimination of individual cowpea seeds to different classes. Considering the short time of image acquisition and limited sample preparation, this stat-of-the art multispectral imaging method and chemometric analysis in classifying seeds could be a valuable tool for on-line classification protocols in cost-effective real-time sorting and grading processes as it provides not only morphological and physical features but also chemical information for the seeds being examined. Implementing image processing algorithms specific for seed quality assessment along with the declining cost and increasing power of computer hardware is very efficient to make the development of such computer-integrated systems more attractive in automatic inspection of seed quality.

Detection of adulteration with duck meat in minced lamb meat by using visible near-infrared hyperspectral imaging

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 (R²_P) 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.