Near-infrared hyperspectral imaging for predicting colour, pH and tenderness of fresh beef

Exploring the potential of hyperspectral imaging for microbial assessment of meat: A review

Food safety is always of paramount importance globally due to the devasting social and economic effects of foodborne disease outbreaks. There is a high consumption rate of meat worldwide, making it an essential protein source in the human diet, hence its microbial safety is of great importance. The food industry stakeholders are always in search of methods that ensure safe food whilst maintaining food quality and excellent sensory attributes. Currently, there are several methods used in microbial food analysis, however, these methods are often time-consuming and do not allow real-time analysis. Considering the recent technological breakthroughs in artificial intelligence and machine learning, it raises the question of whether these advancements could be leveraged within the meat industry to improve turnaround time for microbial assessments. Hyperspectral imaging (HSI) is a highly prospective technology worth exploring for microbial analysis. The rapid, non-destructive method has the potential to be integrated into food production systems and allows foodborne pathogen detection in food samples, thus saving time. Although there has been a substantial increase in research on the utilisation of HSI in food applications over the past years, its use in the microbial assessment of meat is not yet optimal. This review aims to provide a basic understanding of the visible-near infrared HSI system, recent applications in the microbial assessment of meat products, challenges, and possible future applications.

Estimating yolk weight of duck eggs using VIS-NIR Spectroscopy and RGB images and whole egg weights

Duck eggs are widely-consumed food and cooking ingredient. The heavier yolk weight (YW) corresponds to a larger size and greater value. However, there is no nondestructive method available to estimate the weight of the yolk. Accurate weight prediction of duck egg yolks must combine both phenotypic and internal information. In this research, we used Visible-Near Infrared (VIS-NIR) spectroscopy to obtain internal information of duck eggs, and a high-definition camera to capture their phenotypic features. YW was predicted by combining the reduced spectral and RGB image information with the whole egg weight. We also investigated the impact of color and thickness of the duck egg on spectral transmittance (ST), as these factors can influence the extent of ST. The results showed that the spectral curves of duck eggs produced 2 peaks and 1 valley, which may be caused by the dual-frequency absorption of the C-H group and O-H group, and can be used to symbolize the internal information of duck eggs. The ST was somewhat affected by the color and thickness of the duck eggshell. Before modelling, Principal component analysis (PCA) was used to significantly reduce the dimensionality of the RGB image with spectral data. A partial least squares regression (PLSR) model was utilized to fit all the features. The test set yielded a coefficient of determination (R2) of 0.82 and a Root Mean Squared Error (RMSE) of 1.05 g. After removing the eggshell's color and thickness features, the model showed an R2 of 0.79 and an RMSE of 1.11 g. This study demonstrated that the yolk weight of duck eggs can be estimated using VIS-NIR spectroscopy, RGB images and whole egg weight. Furthermore, the effects of shell color and thickness can be neglected.

Smartphone-based digital phenotyping for genome-wide association study of intramuscular fat traits in longissimus dorsi muscle of pigs

Intramuscular fat (IMF) content and distribution significantly contribute to the eating quality of pork. However, the current methods used for measuring these traits are complex, time-consuming and costly. To simplify the measurement process, this study developed a smartphone application (App) called Pork IMF. This App serves as a rapid and portable phenotyping tool for acquiring pork images and extracting the image-based IMF traits through embedded deep-learning algorithms. Utilizing this App, we collected the IMF traits of the longissimus dorsi muscle in a crossbred population of Large White × Tongcheng pigs. Genome-wide association studies detected 13 and 16 SNPs that were significantly associated with IMF content and distribution, respectively, highlighting NR2F2, MCTP2, MTLN, ST3GAL5, NDUFAB1 and PID1 as candidate genes. Our research introduces a user-friendly digital phenotyping technology for quantifying IMF traits and suggests candidate genes and SNPs for genetic improvement of IMF traits in pigs.

Handling the Challenges of Small-Scale Labeled Data and Class Imbalances in Classifying the N and K Statuses of Rubber Leaves Using Hyperspectroscopy Techniques

The nutritional status of rubber trees (Hevea brasiliensis) is inseparable from the production of natural rubber. Nitrogen (N) and potassium (K) levels in rubber leaves are 2 crucial criteria that reflect the nutritional status of the rubber tree. Advanced hyperspectral technology can evaluate N and K statuses in leaves rapidly. However, high bias and uncertain results will be generated when using a small size and imbalance dataset to train a spectral estimaion model. A typical solution of laborious long-term nutrient stress and high-intensive data collection deviates from rapid and flexible advantages of hyperspectral tech. Therefore, a less intensive and streamlined method, remining information from hyperspectral image data, was assessed. From this new perspective, a semisupervised learning (SSL) method and resampling techniques were employed for generating pseudo-labeling data and class rebalancing. Subsequently, a 5-classification spectral model of the N and K statuses of rubber leaves was established. The SSL model based on random forest classifiers and mean sampling techniques yielded optimal classification results both on imbalance/balance dataset (weighted average precision 67.8/78.6%, macro averaged precision 61.2/74.4%, and weighted recall 65.7/78.5% for the N status). All data and code could be viewed on the:Github https://github.com/WeehowTang/SSL-rebalancingtest. Ultimately, we proposed an efficient way to rapidly and accurately monitor the N and K levels in rubber leaves, especially in the scenario of small annotation and imbalance categories ratios.

Potential of fluorescence fingerprints for fish meat authentication: Differences in freshness evaluation in white and dark meat at frozen state

As dark meat has a faster deterioration rate and its unintentional mixing occurs during processing, it is crucial to know the status and freshness indicators of dark meat to ensure fishery product quality. In this method, fluorescence fingerprints (FFs) was applied as a rapid and noninvasive quality authentication method to determine differences between white and dark meat in the evaluation of freshness indicators at frozen state. Spotted mackerel (Scomber australasicus) fish chunks with different postmortem conditions (0-40 h ice stored) were obtained and frozen. A new generation of fluorescence spectrophotometer (F-7100) was used to acquire FFs of the frozen fish chunks (containing white and dark meat). Adenosine triphosphate metabolites and pH were determined in both white and dark meat using their relevant biochemical methods. Higher K-values in dark meat might be attributed to a higher accumulation rate of inosine (HxR) in dark meat than in white meat. The pH decrease rate in white meat was higher than that in dark meat during postmortem ice storage periods of fish. Principal component analysis of FFs spectra demonstrated clear discrimination (PC1 + PC2 = 91.7%) between white and dark meat of frozen fish due to the influence of freshness parameters based on the fluorescence features of fish meat. Furthermore, partial least squares regression validation models revealed that freshness indicators of white meat could be predicted more accurately at the frozen state than those of dark meat. This method could be applied during the processing of fishery products, thereby facilitating quality control activities and making it a promising authentication tool for the fisheries industries.

Rapid and Non-Invasive Assessment of Texture Profile Analysis of Common Carp (Cyprinus carpio L.) Using Hyperspectral Imaging and Machine Learning

Hyperspectral imaging (HSI) has been applied to assess the texture profile analysis (TPA) of processed meat. However, whether the texture profiles of live fish muscle could be assessed using HSI has not been determined. In this study, we evaluated the texture profile of four muscle regions of live common carp by scanning the corresponding skin regions using HSI. We collected skin hyperspectral information from four regions of 387 scaled and live common carp. Eight texture indicators of the muscle corresponding to each skin region were measured. With the skin HSI of live common carp, six machine learning (ML) models were used to predict the muscle texture indicators. Backpropagation artificial neural network (BP-ANN), partial least-square regression (PLSR), and least-square support vector machine (LS-SVM) were identified as the optimal models for predicting the texture parameters of the dorsal (coefficients of determination for prediction (rp) ranged from 0.9191 to 0.9847, and the root-mean-square error for prediction ranged from 0.1070 to 0.3165), pectoral (rp ranged from 0.9033 to 0.9574, and RMSEP ranged from 0.2285 to 0.3930), abdominal (rp ranged from 0.9070 to 0.9776, and RMSEP ranged from 0.1649 to 0.3601), and gluteal (rp ranged from 0.8726 to 0.9768, and RMSEP ranged from 0.1804 to 0.3938) regions. The optimal ML models and skin HSI data were employed to generate visual prediction maps of TPA values in common carp muscles. These results demonstrated that skin HSI and the optimal models can be used to rapidly and accurately determine the texture qualities of different muscle regions in common carp.

Meat color and iridescence: Origin, analysis, and approaches to modulation

Meat color is an important aspect for the meat industry since it strongly determines the consumers' perception of product quality and thereby significantly influences the purchase decision. Emergence of new vegan meat analogs has renewed interest in the fundamental aspects of meat color in order to replicate it. The appearance of meat is based on a complex interplay between the pigment-based meat color from myoglobin and its chemical forms and light scattering from the muscle's microstructure. While myoglobin biochemistry and pigment-based meat color have been extensively studied, research on the physicochemical contribution of light scattering to meat color and the special case of structural colors causing meat iridescence has received only little attention. Former review articles focused mostly on the biochemical or physical mechanisms rather than the interplay between them, in particular the role that structural colors play. While from an economic point of view, meat iridescence might be considered negligible, an enhanced understanding of the underlying mechanisms and the interactions of light with meat microstructures can improve our overall understanding of meat color. Therefore, this review discusses both biochemical and physicochemical aspects of meat color including the origin of structural colors, highlights new color measurement methodologies suitable to investigate color phenomena such as meat iridescence, and finally presents approaches to modulate meat color in terms of base composition, additives, and processing.

SVM Regression to Assess Meat Characteristics of Bísaro Pig Loins Using NIRS Methodology

This study evaluates the ability of the near infrared reflectance spectroscopy (NIRS) to estimate the aW, protein, moisture, ash, fat, collagen, texture, pigments, and WHC in the Longissimus thoracis et lumborum (LTL) of Bísaro pig. Samples (n = 40) of the LTL muscle were minced and scanned in an FT-NIR MasterTM N500 (BÜCHI) over a NIR spectral range of 4000-10,000 cm^-1 with a resolution of 4 cm^-1. The PLS and SVM regression models were developed using the spectra's math treatment, DV1, DV2, MSC, SNV, and SMT (n = 40). PLS models showed acceptable fits (estimation models with RMSE ≤ 0.5% and R² ≥ 0.95) except for the RT variable (RMSE of 0.891% and R² of 0.748). The SVM models presented better overall prediction results than those obtained by PLS, where only the variables pigments and WHC presented estimation models (respectively: RMSE of 0.069 and 0.472%; R² of 0.993 and 0.996; slope of 0.985 ± 0.006 and 0.925 ± 0.006). The results showed NIRs capacity to predict the meat quality traits of Bísaro pig breed in order to guarantee its characterization.

Raw Beef Patty Analysis Using Near-Infrared Hyperspectral Imaging: Identification of Four Patty Categories

South African legislation regulates the classification/labelling and compositional specifications of raw beef patties, to combat processed meat fraud and to protect the consumer. A near-infrared hyperspectral imaging (NIR-HSI) system was investigated as an alternative authentication technique to the current destructive, time-consuming, labour-intensive and expensive methods. Eight hundred beef patties (ca. 100 g) were made and analysed to assess the potential of NIR-HSI to distinguish between the four patty categories (200 patties per category): premium 'ground patty'; regular 'burger patty'; 'value-burger/patty' and the 'econo-burger'/'budget'. Hyperspectral images were acquired with a HySpex SWIR-384 (short-wave infrared) imaging system using the Breeze^® acquisition software, in the wavelength range of 952-2517 nm, after which the data was analysed using image analysis, multivariate techniques and machine learning algorithms. It was possible to distinguish between the four patty categories with accuracies ≥97%, indicating that NIR-HSI offers an accurate and reliable solution for the rapid identification and authentication of processed beef patties. Furthermore, this study has the potential of providing an alternative to the current authentication methods, thus contributing to the authenticity and fair-trade of processed meat products locally and internationally.

Use of Near-Infrared Spectroscopy to Discriminate DFD Beef and Predict Meat Quality Traits in Autochthonous Breeds

The potential of near-infrared reflectance spectroscopy (NIRS) to discriminate Normal and DFD (dark, firm, and dry) beef and predict quality traits in 129 Longissimus thoracis (LT) samples from three Spanish purebreeds, Asturiana de los Valles (AV; n = 50), Rubia Gallega (RG; n = 37), and Retinta (RE; n = 42) was assessed. The results obtained by partial least squares-discriminant analysis (PLS-DA) indicated successful discrimination between Normal and DFD samples of meat from AV and RG (with sensitivity over 93% for both and specificity of 100 and 72%, respectively), while RE and total sample sets showed poorer results. Soft independent modelling of class analogies (SIMCA) showed 100% sensitivity for DFD meat in total, AV, RG, and RE sample sets and over 90% specificity for AV, RG, and RE, while it was very low for the total sample set (19.8%). NIRS quantitative models by partial least squares regression (PLSR) allowed reliable prediction of color parameters (CIE L*, a*, b*, hue, chroma). Results from qualitative and quantitative assays are interesting in terms of early decision making in the meat production chain to avoid economic losses and food waste.

Determination of geographic origins and types of Lindera aggregata samples using a portable short-wave infrared hyperspectral imager

A portable short-wavelength infrared microscope hyperspectral imager (SMHI) combined with machine learning algorithms for the purpose of classifying geographical origins as well as root types of Lindera aggregata is developed. The spectral range of the SMHI system is 1090-1820 nm (5500-9100 cm-1) with spectral and spatial resolutions of 4 nm and 27.3 μm, respectively. Utilizing PCA-RF algorithms, the geographic origin of tuberous roots and leaves from five different origins were classified with accuracies of 97.5% and 97.8%, respectively. In addition, spatial identification of tuberous root and taproot tubers in a mixed sample was done with an accuracy of 98.98%. The accuracy of origin classification and spatial identification are high enough which indicate the significant potential of applying SMHI system into the non-invasive spatial mapping and rapid quality assessment of medicinal herbs.

Tandem mass tag-based proteomics analysis of protein changes in the freezing and thawing cycles of Trachurus murphyi

We investigated the proteome variations in Trachurus murphyi with different cycles of freezing and thawing (FT) under frozen storage. A total of 2,482 proteins were assessed quantitatively, of which 269 proteins were recognized as differential abundance proteins during the second FT cycle until the eighth FT cycle. Bioinformatics analysis on gene ontology (GO) functional annotation and Kyoto Encyclopedia of Genes and Genomes pathway analyses of Differential Analysis of Proteins (DAPs) indicated multiple DAPs engaged with the protein structure, metabolic enzymes, and protein turnover. In addition, some of the observed proteins were probably the underlying markers of protein oxidation (PO). The analysis of PO sites revealed the sites of PO, such as amino adipic semialdehydes, γ-glutamic semialdehydes, and Schiff bases. Bioinformatics analyses demonstrated the involvement of differentially expressed proteins in the Hippo signaling pathway (Ko04390), indicating strong protein degradation with greater numbers of FT cycles under frozen storage. It provides an insight into quality stability from a proteomics quality perspective at the molecular level. The results obtained have deepened our current understandings of the mechanisms that reveal variations in proteomes and quality, as well as help promote quality control of T. murphyi across the cold transportation chain. PRACTICAL APPLICATION: Temperature fluctuation is one of the core issues during frozen food storage and distribution faced by the frozen food industry. Fluctuation may result in microstructural changes, ice recrystallization, and protein change in frozen food products. Tandem mass tag-based methods were adopted to study proteome variations in Trachurus murphyi muscles under different cycles of freezing and thawing under frozen storage conditions in this paper. The results obtained have deepened our current understandings of the mechanisms that reveal variations in proteomes and quality, as well as help promote quality control of T. murphyi across the cold transportation chain.

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.

Meat 4.0: Principles and Applications of Industry 4.0 Technologies in the Meat Industry

Meat 4.0 refers to the application the fourth industrial revolution (Industry 4.0) technologies in the meat sector. Industry 4.0 components, such as robotics, Internet of Things, Big Data, augmented reality, cybersecurity, and blockchain, have recently transformed many industrial and manufacturing sectors, including agri-food sectors, such as the meat industry. The need for digitalised and automated solutions throughout the whole food supply chain has increased remarkably during the COVID-19 pandemic. This review will introduce the concept of Meat 4.0, highlight its main enablers, and provide an updated overview of recent developments and applications of Industry 4.0 innovations and advanced techniques in digital transformation and process automation of the meat industry. A particular focus will be put on the role of Meat 4.0 enablers in meat processing, preservation and analyses of quality, safety and authenticity. Our literature review shows that Industry 4.0 has significant potential to improve the way meat is processed, preserved, and analysed, reduce food waste and loss, develop safe meat products of high quality, and prevent meat fraud. Despite the current challenges, growing literature shows that the meat sector can be highly automated using smart technologies, such as robots and smart sensors based on spectroscopy and imaging technology.

Microscopy is better in color: development of a streamlined spectral light path for real-time multiplex fluorescence microscopy

Spectroscopic image data has provided molecular discrimination for numerous fields including: remote sensing, food safety and biomedical imaging. Despite the various technologies for acquiring spectral data, there remains a trade-off when acquiring data. Typically, spectral imaging either requires long acquisition times to collect an image stack with high spectral specificity or acquisition times are shortened at the expense of fewer spectral bands or reduced spatial sampling. Hence, new spectral imaging microscope platforms are needed to help mitigate these limitations. Fluorescence excitation-scanning spectral imaging is one such new technology, which allows more of the emitted signal to be detected than comparable emission-scanning spectral imaging systems. Here, we have developed a new optical geometry that provides spectral illumination for use in excitation-scanning spectral imaging microscope systems. This was accomplished using a wavelength-specific LED array to acquire spectral image data. Feasibility of the LED-based spectral illuminator was evaluated through simulation and benchtop testing and assessment of imaging performance when integrated with a widefield fluorescence microscope. Ray tracing simulations (TracePro) were used to determine optimal optical component selection and geometry. Spectral imaging feasibility was evaluated using a series of 6-label fluorescent slides. The LED-based system response was compared to a previously tested thin-film tunable filter (TFTF)-based system. Spectral unmixing successfully discriminated all fluorescent components in spectral image data acquired from both the LED and TFTF systems. Therefore, the LED-based spectral illuminator provided spectral image data sets with comparable information content so as to allow identification of each fluorescent component. These results provide proof-of-principle demonstration of the ability to combine output from many discrete wavelength LED sources using a double-mirror (Cassegrain style) optical configuration that can be further modified to allow for high speed, video-rate spectral image acquisition. Real-time spectral fluorescence microscopy would allow monitoring of rapid cell signaling processes (i.e., Ca²⁺ and other second messenger signaling) and has potential to be translated to clinical imaging platforms.

Hyperspectral imaging (HSI) for meat quality evaluation across the supply chain: Current and future trends

Meat products are particularly plagued by safety problems because of their complicated structure, various production processes and complex supply chains. Rapid and non-invasive analytical methods to evaluate meat quality have become a priority for the industry over the conventional chemical methods. To achieve rapid analysis of safety and quality parameters of meat products, hyperspectral imaging (HSI) is now widely applied in research studies for detecting the various components of different meat products, but its application in meat production and supply chain integrity as a quality control (QC) solution is still ambiguous. This review presents the fresh look at the current states of HSI research as both the scope and the applicability of the HSI in the meat quality evaluation expanded. The future application scenarios of HSI in the supply chain and the future development of HSI hardware and software are also discussed, by which HSI technology has the potential to enable large scale meat product testing. With a fully adapted for factory setting HSI, the inspection coverage can reliably identify the chemical properties of meat products. With the introduction of Food Industry 4.0, HSI advances can change the meat industry to become from reactive to predictive when facing meat safety issues.

•

HSI has shown promising early signs in the non-destructive analysis of meat quality and safety.

•

Hyperspectral imaging (HSI) is now widely applied in research studies for different meat products with the help of machine learning methods.

•

With a fully adapted factory setting and robust machine learning of HSI, the inspection coverage can reach 100% of the target meat.

•

HSI can change the meat industry to become from reactive to predictive when facing issues, this will be translated into fewer recalls, less meat fraud, and less waste.

Recent technology for food and beverage quality assessment: a review

Food and beverage assessment is an evaluation method used to measure the strengths and weaknesses of a food and beverage system to make improvements. These assessments had become crucial, especially in the issues of adulteration, replacement, and contamination that happened in artificial adjustment relating to the quality, weight and volume. Thus, this review will examine and describe features recently applied in image, odour, taste and electromagnetic, relevant to the food and beverages assessment. This review will also compare and discuss each technique and provides suggestions based on the current technology. This review will deliberate technology integration and the involvement of deep learning to enable several types of current technologies, such as imaging, odour and taste senses, and electromagnetic sensing, to be used in food evaluation applications for inspection and packaging.

Programmable black phosphorus image sensor for broadband optoelectronic edge computing

Image sensors with internal computing capability enable in-sensor computing that can significantly reduce the communication latency and power consumption for machine vision in distributed systems and robotics. Two-dimensional semiconductors have many advantages in realizing such intelligent vision sensors because of their tunable electrical and optical properties and amenability for heterogeneous integration. Here, we report a multifunctional infrared image sensor based on an array of black phosphorous programmable phototransistors (bP-PPT). By controlling the stored charges in the gate dielectric layers electrically and optically, the bP-PPT’s electrical conductance and photoresponsivity can be locally or remotely programmed with 5-bit precision to implement an in-sensor convolutional neural network (CNN). The sensor array can receive optical images transmitted over a broad spectral range in the infrared and perform inference computation to process and recognize the images with 92% accuracy. The demonstrated bP image sensor array can be scaled up to build a more complex vision-sensory neural network, which will find many promising applications for distributed and remote multispectral sensing.

2D materials represent a promising platform for machine vision and edge computing applications, although usually limited to ultraviolet and visible wavelengths. Here, the authors report the realization of a programmable image sensor based on black phosphorus, implementing multispectral imaging and analog in-memory computing functionalities in the near- to mid-infrared range.

Multispectral Wavebands Selection for the Detection of Potential Foreign Materials in Fresh-Cut Vegetables

Ensuring the quality of fresh-cut vegetables is the greatest challenge for the food industry and is equally as important to consumers (and their health). Several investigations have proven the necessity of advanced technology for detecting foreign materials (FMs) in fresh-cut vegetables. In this study, the possibility of using near infrared spectral analysis as a potential technique was investigated to identify various types of FMs in seven common fresh-cut vegetables by selecting important wavebands. Various waveband selection methods, such as the weighted regression coefficient (WRC), variable importance in projection (VIP), sequential feature selection (SFS), successive projection algorithm (SPA), and interval PLS (iPLS), were used to investigate the optimal multispectral wavebands to classify the FMs and vegetables. The application of selected wavebands was further tested using NIR imaging, and the results showed good potentiality by identifying 99 out of 107 FMs. The results indicate the high applicability of the multispectral NIR imaging technique to detect FMs in fresh-cut vegetables for industrial application.

Using Ensemble-Based Systems with Near-Infrared Hyperspectral Data to Estimate Seasonal Snowpack Density

Estimating the seasonal density of the snowpack has many financial and environmental benefits. Rapid assessment and daily monitoring of its evolution are therefore key to effective prevention. Traditionally, the physical characteristics of snow are measured directly in the field, which involves high costs and personnel mobilization. Hyperspectral imaging is a reliable and efficient technique to study and evaluate this physical property. The spectral reflectance of snow is partly defined by changes in its physical properties, particularly in the Near infrared (NIR) part of the spectrum. Recently, a hybrid snow density estimation model allowing retrieval of density from NIR hyperspectral data was developed, based on an a priori classification of snow samples. However, in order to obtain optimal density estimates with the Hybrid model (HM), the sources of classification and estimation error must be controlled. Following the same principle as the HM, an Ensemble-based system (EBS) was developed. This model reduces the number of misclassification errors produced by the HM. The general concept of EBS algorithms is based on the principle that obtaining more opinions before making a decision is part of human nature, especially when economic and environmental benefits are at stake. This approach has helped to reduce the risk of classification and estimation errors and to develop more robust density results. One hundred and fourteen snow samples collected during three winters (2018–2020) were used to calibrate and validate the EBS. The performance of the EBS was validated using an independent database and the results were satisfactory (R2 = 0.90, RMSE = 44.45 kg m−3, BIAS = 3.87 kg m−3 and NASH = 0.89).

Estimating Snowpack Density from Near-Infrared Spectral Reflectance Using a Hybrid Model

Improving the estimation of snow density is a key task in current snow research. Characterization of the variability of density in time and space is essential for the estimation of water equivalent, hydroelectric power production, assessment of natural hazards (avalanches, floods, etc.). Hyperspectral imaging is proving to be a promising and reliable tool for monitoring and estimating this physical property. Indeed, the spectral reflectance of snow is partly controlled by changes in its physical properties, particularly in the near-infrared (NIR) part of the spectrum. For this purpose, several models have been designed to estimate snow density from spectral information. However, none has yet achieved significant performance. One of the major difficulties is that the relationship between snow density and spectral reflectance is non-bijective (surjective). Indeed, several reflectance amplitudes can be associated with the same density and vice versa, so the correlation between density and spectral reflectance can be very poor. To resolve this issue, a hybrid snow density estimation model based on spectral data is proposed in this work. The principle behind this model is to classify the snow density prior to its estimation by means of a specific estimator corresponding to a predetermined snow density class. These additional steps eliminate the surjective relation by converting it into three bijective relations between density and spectral reflectance. The calibration step showed that the densities included within the three classes are sensitive to different spectral regions, with R2 > 0.80. The results of the cross-validation for the specific estimators were also satisfactory with R2 > 0.78 and RMSE < 36.36 kg m−3. The overall performance of the hybrid model (HM), when tested with independent data, demonstrated the effectiveness of using proximal NIR hyperspectral imagery to estimate snow density (R2 = NASH = 0.93).

Prediction of moisture content in steamed and dried purple sweet potato using hyperspectral imaging analysis

Partial least squares regression (PLSR) modeling was performed to predict the moisture content in steamed, dried purple sweet potato based on spectral data obtained from hyperspectral imaging analysis. The PLSR model with a combination of multiplicative scatter correction, Savitzky-Golay, and first derivative exhibited the highest accuracy (R_P ² = 0.9754). The wavelengths found that strongly affected the PLSR model were 961.12, 1065.50, 1083.93, 1173.23, and 1233.89 nm. These wavelengths were associated with the O-H second overtone and the second overtone of C-H, C-H₂, and C-H₃. When PLSR modeling was performed using these selected wavelengths, the prediction accuracy of the PLSR model exhibited high accuracy (R_P ² = 0.9521). Therefore, the moisture content could be predicted with high accuracy using only five wavelengths rather than the full spectrum.

Artificial Intelligence Empowered Multispectral Vision Based System for Non-Contact Monitoring of Large Yellow Croaker (Larimichthys crocea) Fillets

A non-contact method was proposed to monitor the freshness (based on TVB-N and TBA values) of large yellow croaker fillets (Larimichthys crocea) by using a visible and near-infrared hyperspectral imaging system (400-1000 nm). In this work, the quantitative calibration models were built by using feed-forward neural networks (FNN) and partial least squares regression (PLSR). In addition, it was established that using a regression coefficient on the data can be further compressed by selecting optimal wavelengths (35 for TVB-N and 18 for TBA). The results validated that FNN has higher prediction accuracies than PLSR for both cases using full and selected reflectance spectra. Moreover, our FNN based model has showcased excellent performance even with selected reflectance spectra with rp = 0.978, R2p = 0.981, and RMSEP = 2.292 for TVB-N, and rp = 0.957, R2p = 0.916, and RMSEP = 0.341 for TBA, respectively. This optimal FNN model was then utilized for pixel-wise visualization maps of TVB-N and TBA contents in fillets.

Impact of drying temperature and salt pre-treatments on drying behavior and instrumental color and investigations on spectral product monitoring during drying of beef slices

Drying behavior and instrumental color development of beef slices untreated or pretreated with salt or salt and vinegar solutions were monitored by determining the moisture content and the color change by measuring CIELAB values during drying at 50, 60, and 70 °C. Time-series hyperspectral imaging (400-1000 nm) was applied with regard to the development of non-invasive measurement systems based on robust models to predict moisture and color independent of the pre-treatment and drying temperature. Samples pretreated with salt dried the slowest which became more prominent at increasing drying temperatures and the least color change (∆E = 23) was observed at 60 °C drying temperature. Robust prediction models for moisture content and CIELAB values irrespective of pre-treatment and processing conditions were developed successfully and improved by wavelengths selection with high R² (0.94-0.98) and low RMSEP (1.05-5.22) which will support the future development of simple and cost-effective applications regarding non-invasive product monitoring systems for beef drying processes.

Comparison of smartphone and lab-grade NIR spectrometers to measure chemical composition of lamb and beef

Context Near-infrared reflectance spectroscopy (NIRS) has been extensively investigated for non-destructive and rapid determination of pH and chemical composition of meat including water, crude protein, intramuscular fat (IMF) and stable isotopes. Smaller, cheaper NIRS sensors that connect to a smartphone could enhance the accessibility and uptake of this technology by consumers. However, the limited wavelength range of these sensors could restrict the accuracy of predictions compared with benchtop laboratory NIRS models. Aims To compare the precision and accuracy metrics of predicting pH, water, crude protein and IMF of three sample presentations and two sensors. Methods Fresh intact (FI) store-bought beef and lamb steak samples (n = 43) were ground and freeze-dried (FD), and then oven-dried to create freeze-dried oven-dried (FDOD) samples. All three forms of sample presentation (FI, FD, FDOD) were scanned using the smartphone and benchtop NIRS sensors. Key results The IMF was the best predicted trait in FD and FDOD forms by the smartphone NIRS (R2 &gt;0.75; RPD &gt;1.40) with limited differences between the two sensors. However, predictions on FI meat were poorer for all traits regardless of the NIRS scanner used (R2 ≤ 0.67; RPD ≤ 1.58) and not suitable for use in research or industry. Conclusion The smartphone NIRS sensor showed accuracy and precision comparable to benchtop NIRS to predict meat composition. However, these preliminary results found that neither of the two sensors reliably predicted quality attributes for industry or consumer applications. Implications Miniaturised NIRS sensors connected to smartphones could provide a practical solution to measure some meat quality attributes such as IMF, but the accuracy depends on sample presentation.

Collagen, intramuscular fat and proteolysis affect Warner-Bratzler shear-force of muscles from Bos taurus breed types differently at weaning, after backgrounding on pasture, and after feedlotting

Context The texture of beef is highly important for the eating experience, and there is a continued interest in understanding the biochemical basis for the variation in texture between cattle and their meat cuts in order to improve and minimise variation in tenderness due to production and processing factors. Aims The present study aimed to investigate the impact of characteristics of meat on Warner-Bratzler shear-force (WBSF) as an indicator of texture of beef as affected by breed type, age/feeding phase, and muscle. Methods Seventy-five steers of three breed types (Angus, Hereford and Wagyu × Angus) were slaughtered after weaning 6 months old (n = 15), after backgrounding 17 months old (n = 30) and after feedlotting 25 months old (n = 30). At slaughter three muscles (M. supraspinatus, M. semitendinosus and M. longissimus lumborum) were sampled from each steer, and pH, intramuscular fat and collagen content, sarcomere length, and proteolysis (desmin degradation) were measured and used to explain the variation in WBSF after 7 and 14 days of aging. Key results Meat from Hereford and Angus steers had higher WBSF after 7 days of aging compared with Wagyu × Angus steers, but after 14 days of aging there was only a difference between Hereford and Wagyu × Angus in the M. supraspinatus and M. semitendinosus. The WBSF of the young weaned steers and steers slaughtered after backgrounding were dependent on the degree of proteolysis in the muscles, whereas for steers slaughtered after feedlotting the content of collagen was more important for the WBSF. The amount of intramuscular fat had a significant impact on the differences in WBSF within the specific muscle studied. In contrast to the general dogma that WBSF increase with age, WBSF decreased in M. semitendinosus and M. longissimus lumborum from the weaned 6-month-old steers to the 25-month-old steers finished in feed-lot, whereas in M. supraspinatus the older feed-lot finished steers had a higher WBSF. Conclusion The factors contributing to the Warner-Bratzler shear force of beef depends on the age/feeding phase of the animal and the muscle and less on the breed type. Implications Optimisation of texture in beef through breeding and production should address different traits dependent on the age/feeding phase of the slaughter animal.

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.

Rapid noninvasive monitoring of freshness variation in frozen shrimp using multidimensional fluorescence imaging coupled with chemometrics

Shrimp is one of the most delicious and popular food commodities worldwide due to its exceptional taste and characteristics. Freshness is considered as a key factor for shrimp consumers because freshness has a significant relationship with taste and shelf-life of shrimp. However, post-mortem metabolism of shrimp differs from that of fish as they are highly susceptible to post-harvest quality loss, and it is hard to distinguish the freshness variation of shrimp at frozen state instantly. Thus, instant monitoring of frozen shrimp freshness is challenging for the seafood and aquaculture industries and a reliable, expeditious, and noninvasive technique to estimate shrimp quality is in high demand. Accordingly, this study aimed to visualize changes in post-mortem freshness of frozen shrimp using multidimensional fluorescence imaging. Live coonstripe shrimp (Pandalus hypsinotus) were harvested and instantly killed by beheading, cooled on ice for 0, 6, 24, 48, 72 and 96 h (n = 8), followed by processing into frozen peeled deveined shrimp product and stored at -60 °C. 50% of frozen shrimp were analyzed for excitation-emission matrix (EEM), ATP-related compounds, and pH using a fiber optic supported fluorescence spectrophotometer (F-7100), high performance liquid chromatography (HPLC) and pH meter, respectively at each time point (n = 4). Then, fluorescence images were obtained from the remaining 50% of frozen shrimp (n = 4) by computer vision method equipped with a charge-coupled device (CCD) camera, MAX-303 xenon light source for an excitation light (Ex. 330 nm), and an automatic filter changer for emission band-pass filters (Em. 380-610 nm at 10 nm intervals). Chemical analysis of frozen shrimp revealed that K-value and pH of shrimp increased from 1.61 to 66.56% and 6.49-7.31, respectively, during storage on ice. Repeated partial least squares regression (PLSR) models of EEM for K-value prediction suggested an efficient excitation wavelength (330 nm) and its corresponding emission wavelengths (380-610 nm) to produce fluorescence images. Spatial-temporal changes of K-value and pH were visualized successfully in frozen shrimp by fluorescence imaging. K-value visualization was then validated effectively using another group of frozen shrimp (0-72 h ice stored) with different killing method (super chilling) and the prediction accuracy was R² = 0.80. This novel approach using a CCD camera coupled with EEM provides a state-of-the-art authentication method for practical assessment of frozen seafood freshness.

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.

Myoglobin-Based Classification of Minced Meat Using Hyperspectral Imaging

Minced meat substitution is one of the most common frauds which not only affects consumer health but impacts their lifestyles and religious customs as well. A number of methods have been proposed to overcome these frauds; however, these mostly rely on laboratory measures and are often subject to human error. Therefore, this study proposes novel hyperspectral imaging (400–1000 nm) based non-destructive isos-bestic myoglobin (Mb) spectral features for minced meat classification. A total of 60 minced meat spectral cubes were pre-processed using true-color image formulation to extract regions of interest, which were further normalized using the Savitzky–Golay filtering technique. The proposed pipeline outperformed several state-of-the-art methods by achieving an average accuracy of 88.88%.

Prediction of quality traits and grades of intact chicken breast fillets by hyperspectral imaging

1. In this study, hyperspectral imaging was evaluated for its usefulness to predict quality traits and grading of intact chicken breast fillets. 2. Lightness of colour (L*) and pH of the fillets were measured as quality traits, and samples were then selected and graded to three different quality categories, i.e., dark, firm and dry (DFD), normal (NORM), and pale, soft and exudative (PSE) based on these two quality traits. Based on the prediction performance of full wavelength partial least square regression (PLSR) models, the spectral range of visible and near-infrared (Vis-NIR) was more suitable for the evaluation of quality traits and grading than the range of near-infrared (NIR). Key wavelengths of each quality trait and grade value were selected by the regression coefficient (RC) method. 3. The new key wavelength PLSR models showed good predictive performances (Rp = 0.85 and RMSEp = 2.18 for L*, Rp = 0.84, and RMSEp = 0.13 for pH, and Rp = 0.80 and RMSEp = 0.44 for quality grading). The classification accuracy for grades was 85.71% (calibration set) and 81.82% (prediction set), respectively. Finally, distribution maps showed that quality traits and grades of samples were able to be visualised. 4. These results suggested that hyperspectral imaging has the potential for quality prediction of fresh chicken meat.

Wavelength Selection FOR Rapid Identification of Different Particle Size Fractions of Milk Powder Using Hyperspectral Imaging

Hyperspectral imaging (HSI) in the spectral range of 400-1000 nm was tested to differentiate three different particle size fractions of milk powder. Partial least squares discriminant analysis (PLS-DA) was performed to observe the relationship of spectral data and particle size information for various samples of instant milk powder. The PLS-DA model on full wavelengths successfully classified the three fractions of milk powder with a coefficient of prediction 0.943. Principal component analysis (PCA) identified each of the milk powder fractions as separate clusters across the first two principal components (PC₁ and PC₂) and five characteristic wavelengths were recognised by the loading plot of the first three principal components. Weighted regression coefficient (WRC) analysis of the partial least squares model identified 11 important wavelengths. Simplified PLS-DA models were developed from two sets of reduced wavelengths selected by PCA and WRC and showed better performance with predictive correlation coefficients (R_p²) of 0.962 and 0.979, respectively, while PLS-DA with complete spectrum had R_p² of 0.943. Similarly, classification accuracy of PLS-DA was improved to 92.2% for WRC based predictive model. Calculation time was also reduced to 2.1 and 2.8 s for PCA and WRC based simplified PLS-DA models in comparison to the complete spectrum model that was taking 32.2 s on average to predict the classification of milk powder samples. These results demonstrated that HSI with appropriate data analysis methods could become a potential analyser for non-invasive testing of milk powder in the future.

Application of optical technologies in the US poultry slaughter facilities for the detection of poultry carcase condemnation

1. This article reviews the studies on optical technologies for automated poultry carcase inspection, discusses challenges and potential solutions in their real-time applications in poultry slaughter facilities. 2. Over the past few decades, extensive research has been underway to develop an optical technology-based machine vision system for automated inspection of poultry carcases and viscera. Such an automated technology will not only aid in carcase inspection to maximise food safety, but it will also support the U.S. New Poultry Inspection System's aim to foster innovation in poultry processing as well as increase line speed. 3. Many earlier studies based on visible and near-infrared spectroscopy showed promise, but could not be implemented successfully in an on-line poultry processing plant. Currently, multi- and hyper-spectral imaging-based machine vision systems have shown promising outcomes. 5. The critical hurdles for real-time application of automated imaging technology in poultry carcase inspection include high-speed processing lines, slaughter facilities environment and variation in broiler rearing practices. Therefore, further improvement in imaging and machine vision technologies based on physiochemical properties on poultry carcases, the establishment of more technology friendly inspection station, and an integrated data management for different rearing practices are essential to overcome those hurdles.

Research in Beef Tenderness and Palatability in the Era of Big Data

For decades, research has focused on predicting beef palatability using muscle biochemical traits, and various biomarkers. In these approaches, a precise definition of the variable to predict (tenderness assessed by panelists, untrained consumers, or shear force), and repeatability of the measurements are crucial for creating significant data resources for the derivation of robust predictive models, and rigorous validation testing. This “big data” approach also requires careful definition of traits and transparent principles for data sharing and management. As in other fields, meat science researchers should improve the Findability, Accessibility, Interoperability, and Reuse of data (known as the FAIR principles). Furthermore, with the rapid evolution of new measurement technologies, the traits that they measure must be consistently described, enhancing our ability to integrate these new measurements into existing description systems. For beef, strategic choices have been made in order to consider real consumers’ expectations, not well estimated correctly by lab approaches. This strategy has been successfully developed in Australia, which set up the “Meat Standards Australia” grading scheme, now partly adopted by the beef industry. The ambitions of the International Meat Research 3G Foundation is to develop beef ontology, to set up an international database with a huge number of consumers’ scores related to beef palatability and collected according to standard protocols. The foundation also aims to support the beef industry by offering an international predictive model of beef palatability, flexible enough to take into account any local livestock characteristics or regional consumer specificity. This approach is supported by the United Nations Economic Commission for Europe (UNECE), which is promoting development of regulations and norms, technical cooperation and exchange of best expertise and practices. This will substantially improve the transparency of data flow and price signaling between all participants of the value chain, from beef producers through to consumers at retail.

Determination of pectin content in orange peels by near infrared hyperspectral imaging

Pectin has several purposes in the food and pharmaceutical industry making its quantification important for further extraction. Current techniques for pectin quantification require its extraction using chemicals and producing residues. Determination of pectin content in orange peels was investigated using near infrared hyperspectral imaging (NIR-HSI). Hyperspectral images from orange peel (140 samples) with different amounts of pectin were acquired in the range of 900-2500 nm, and the spectra was used for calibration models using multivariate statistical analyses. Principal component analysis (PCA) and linear discriminant analysis (LDA) showed better results considering three groups: low (0-5%), intermediate (10-40%) and high (50-100%) pectin content. Partial least squares regression (PLSR) models based on full spectra showed higher precision (R2 > 0.93) than those based on few selected wavelengths (R2 between 0.92 and 0.94). The results demonstrate the potential of NIR-HSI to quantify pectin content in orange peels, providing a valuable technique for orange producers and processing industries.