Multi-task convolutional neural network for simultaneous monitoring of lipid and protein oxidative damage in frozen-thawed pork using hyperspectral imaging

Rapid non-destructive monitoring and quality assessment of the fumigation process of Shanxi aged vinegar based on Vis-NIR hyperspectral imaging combined with multiple chemometric algorithms

The quality of the grains during the fumigation process can significantly affect the flavour and nutritional value of Shanxi aged vinegar (SAV). Hyperspectral imaging (HSI) was used to monitor the extent of fumigated grains, and it was combined with chemometrics to quantitatively predict three key physicochemical constituents: moisture content (MC), total acid (TA) and amino acid nitrogen (AAN). The noise reduction effects of five spectral preprocessing methods were compared, followed by the screening of optimal wavelengths using competitive adaptive reweighted sampling. Support vector machine classification was employed to establish a model for discriminating fumigated grains, and the best recognition accuracy reached 100%. Furthermore, the results of partial least squares regression slightly outperformed support vector machine regression, with correlation coefficient for prediction (Rp) of 0.9697, 0.9716, and 0.9098 for MC, TA, and AAN, respectively. The study demonstrates that HSI can be employed for rapid non-destructive monitoring and quality assessment of the fumigation process in SAV.

Nondestructive detection of lead content in oilseed rape leaves under silicon action using hyperspectral image

This study explored the feasibility of employing hyperspectral imaging (HSI) technology to quantitatively assess the effect of silicon (Si) on lead (Pb) content in oilseed rape leaves. Aiming at the defects of hyperspectral data with high dimension and redundant information, this paper proposed two improved feature wavelength extraction algorithms, repetitive interval combination optimization (RICO) and interval combination optimization (ICO) combined with stepwise regression (ICO-SR). The entire oilseed rape leaves were taken as the region of interest (ROI) to extract the visible near-infrared hyperspectral data within the 400.89-1002.19 nm range. In data processing, Savitzky-Golay (SG) smoothing, detrending (DT), and multiple scatter correction (MSC) were utilized for spectral data preprocessing, while recursive feature elimination (RFE), iteratively variable subset optimization (IVSO), ICO, and the two enhanced algorithms were employed to identify characteristic wavelengths. Subsequently, based on the spectral data of preprocessing and feature extraction, partial least squares regression (PLSR) and support vector regression (SVR) methods were used to construct various Pb content prediction models in oilseed rape leaves, with a comparison and analysis of each model performance. The results indicated that the two improved algorithms were more efficient in extracting representative spectral information than conventional methods, and the performance of SVR models was better than PLSR models. Finally, to further improve the prediction accuracy and robustness of the SVR models, the whale optimization algorithm (WOA) was introduced to optimize their parameters. The findings demonstrated that the MSC-RICO-WOA-SVR model achieved the best comprehensive performance, with Rp2 of 0.9436, RMSEP of 0.0501 mg/kg, and RPD of 3.4651. The results further confirmed the great potential of HSI combined with feature extraction algorithms to evaluate the effectiveness of Si in alleviating Pb stress in oilseed rape and provided a theoretical basis for determining the appropriate amount of Si application to alleviate Pb pollution in oilseed rape.

An effective deep learning fusion method for predicting the TVB-N and TVC contents of chicken breasts using dual hyperspectral imaging systems

Total volatile basic nitrogen (TVB-N) and total viable count (TVC) are important freshness indicators of meat. Hyperspectral imaging combined with chemometrics has been proven to be effective in meat detection. However, a challenge with chemometrics is the lack of a universally applicable processing combination, requiring trial-and-error experiments with different datasets. This study proposes an end-to-end deep learning model, pyramid attention features fusion model (PAFFM), integrating CNN, attention mechanism and pyramid structure. PAFFM fuses the raw visible and near-infrared range (VNIR) and shortwave near-infrared range (SWIR) spectral data for predicting TVB-N and TVC in chicken breasts. Compared with the CNN and chemometric models, PAFFM obtains excellent results without a complicated processing combinatorial optimization process. Important wavelengths that contributed significantly to PAFFM performance are visualized and interpreted. This study offers valuable references and technical support for the market application of spectral detection, benefiting related research and practical fields.

Simultaneous determination of pigments of spinach (Spinacia oleracea L.) leaf for quality inspection using hyperspectral imaging and multi-task deep learning regression approaches

Rapid and accurate determination of pigment content is important for quality inspection of spinach leaves during storage. This study aimed to use hyperspectral imaging at two spectral ranges (visible/near-infrared, VNIR: 400-1000 nm; NIR: 900-1700 nm) to simultaneously determine the pigment (chlorophyll a, chlorophyll b, total chlorophyll, and carotenoids) content in spinach stored at different durations and conditions (unpackaged and packaged). Partial least squares (PLS), back propagation neural network (BPNN) and convolutional neural network (CNN) were used to establish single-task and multi-task regression models. Single-task CNN (STCNN) models and multi-task CNN (MTCNN) models obtained better performances than the other models. The models using VNIR spectra were superior to those using NIR spectra. The overall results indicated that hyperspectral imaging with multi-task learning could predict the quality attributes of spinach simultaneously for spinach quality inspection under various storage conditions. This research will guide food quality inspection by simultaneously inspecting multiple quality attributes.

Rapid Non-Destructive Detection Technology in the Field of Meat Tenderness: A Review

Traditionally, tenderness has been assessed through shear force testing, which is inherently destructive, the accuracy is easily affected, and it results in considerable sample wastage. Although this technology has some drawbacks, it is still the most effective detection method currently available. In light of these drawbacks, non-destructive testing techniques have emerged as a preferred alternative, promising greater accuracy, efficiency, and convenience without compromising the integrity of the samples. This paper delves into applying five advanced non-destructive testing technologies in the realm of meat tenderness assessment. These include near-infrared spectroscopy, hyperspectral imaging, Raman spectroscopy, airflow optical fusion detection, and nuclear magnetic resonance detection. Each technology is scrutinized for its respective strengths and limitations, providing a comprehensive overview of their current utility and potential for future development. Moreover, the integration of these techniques with the latest advancements in artificial intelligence (AI) technology is explored. The fusion of AI with non-destructive testing offers a promising avenue for the development of more sophisticated, rapid, and intelligent systems for meat tenderness evaluation. This integration is anticipated to significantly enhance the efficiency and accuracy of the quality assessment in the meat industry, ensuring a higher standard of safety and nutritional value for consumers. The paper concludes with a set of technical recommendations to guide the future direction of non-destructive, AI-enhanced meat tenderness detection.

Discrimination among Fresh, Frozen-Stored and Frozen-Thawed Beef Cuts by Hyperspectral Imaging

The detection of the storage state of frozen meat, especially meat frozen-thawed several times, has always been important for food safety inspections. Hyperspectral imaging (HSI) is widely applied to detect the freshness and quality of meat or meat products. This study investigated the feasibility of the low-cost HSI system, combined with the chemometrics method, to classify beef cuts among fresh (F), frozen-stored (F-S), frozen-thawed three times (F-T-3) and frozen-thawed five times (F-T-5). A compact, low-cost HSI system was designed and calibrated for beef sample measurement. The classification model was developed for meat analysis with a method to distinguish fat and muscle, a CARS algorithm to extract the optimal wavelength subset and three classifiers to identify each beef cut among different freezing processes. The results demonstrated that classification models based on feature variables extracted from differentiated tissue spectra achieved better performances, with ACCs of 92.75% for PLS-DA, 97.83% for SVM and 95.03% for BP-ANN. A visualization map was proposed to provide detailed information about the changes in freshness of beef cuts after freeze-thawing. Furthermore, this study demonstrated the potential of implementing a reasonably priced HSI system in the food industry.

Study on the aging status of insulators based on hyperspectral imaging technology

The acidic environment is one of the main factors leading to the aging of silicone rubber (SiR) insulators. Aging can reduce the surface hydrophobicity and pollution flashover resistance of insulators, threatening the safe and stable operation of the power grid. Therefore, evaluating the aging state of insulators is essential to prevent flashover accidents on the transmission line. This paper is based on an optical hyperspectral imaging (HSI) technology for pixel-level assessment of insulator aging status. Firstly, the SiR samples were artificially aged in three typical acidic solutions with different concentrations of HNO3, H2SO4, and HCl, and six aging grades of SiR samples were prepared. The HSI of SiR at each aging grade was extracted using a hyperspectral imager. To reduce the calculation complexity and eliminate the interference of useless information in the band, this paper proposes a joint random forest- principal component analysis (RF-PCA) dimensionality reduction method to reduce the original 256-dimensional hyperspectral data to 7 dimensions. Finally, to capture local features in hyperspectral images more effectively and retain the most significant information of the spectral lines, a convolutional neural network (CNN) was used to build a classification model for pixel-level assessment of the SiR's aging state of and visual prediction of insulators' defects. The research method in this paper provides an important guarantee for the timely detection of safety hazards in the power grid.

Wavelength selection method for near-infrared spectroscopy based on Max-Relevance Min-Redundancy

Near-infrared spectroscopy (NIRS) is a rapid, nondestructive analytical technique utilized in various fields. However, the NIR data, which consists of hundreds of dimensions, may exhibit considerable duplication in the spectrum information. This redundancy might impair modeling effectiveness. As a result, feature selection on the spectral data becomes critical. The Max-Relevance Min-Redundancy (mRMR) method stands out among the different feature selection techniques for dimensional reduction. The approach depends on mutual information (MI) between random variables as the basis for feature selection and is unaffected by modeling methods. However, it is necessary to clarify the benefits of the maximum correlation minimal redundancy algorithm in the context of near-infrared spectral feature selection, as well as its adaptability to various modeling methods. This research focuses on the NIR spectral dataset of maize germination rate, and the mRMR method is utilized to select spectral features. Based on the preceding foundation, we create models for Support Vector Regression, Gaussian Process Regression, Random Forest, and Neural Networks. The experimental findings demonstrate that, among the feature selection methods employed in this paper, the Max-Relevance Min-Redundancy algorithm outperforms others regarding the corn germination rate dataset.

Application of Hyperspectral Imaging as a Nondestructive Technology for Identifying Tomato Maturity and Quantitatively Predicting Lycopene Content

Maturity is a crucial indicator in assessing the quality of tomatoes, and it is closely related to lycopene content. Using hyperspectral imaging, this study aimed to monitor tomato maturity and predict its lycopene content at different maturity stages. Standard normal variable (SNV) transformation was applied to preprocess the hyperspectral data. Then, using competitive adaptive reweighted sampling (CARS), the characteristic wavelengths were selected to simplify the calibration models. Based on the full and characteristic wavelengths, a support vector classifier (SVC) model was developed to determine tomato maturity qualitatively. The results demonstrated that the classification accuracy using the characteristic wavelength led to the obtention of better results with an accuracy of 95.83%. In addition, the support vector regression (SVR) and partial least squares regression (PLSR) models were utilized to predict lycopene content. With a coefficient of determination for prediction (R2P) of 0.9652 and a root mean square error for prediction (RMSEP) of 0.0166 mg/kg, the SVR model exhibited the best quantitative prediction capacity based on the characteristic wavelengths. Following this, a visual distribution map was created to evaluate the lycopene content in tomato fruit intuitively. The results demonstrated the viability of hyperspectral imaging for detecting tomato maturity and quantitatively predicting the lycopene content during storage.

Non-Destructive Detection of the Freshness of Air-Modified Mutton Based on Near-Infrared Spectroscopy

Monitoring and identifying the freshness levels of meat holds significant importance in the field of food safety as it directly relates to human dietary safety. Traditional packaging methods for lamb meat quality assessment present issues such as cumbersome operations and irreversible damage. This research proposes a quality assessment method for modified atmosphere packaging lamb meat using near-infrared spectroscopy and multi-parameter fusion. Fresh lamb meat quality is taken as the research subject, comparing various physicochemical indicators and near-infrared spectroscopic information under different temperatures (4 °C and 10 °C) and different modified atmosphere packaging combinations. Through precision parameter comparison, rebound and TVB-N values are selected as the modeling parameters. Six spectral preprocessing methods (multi-scatter calibration, MSC; standard normal variate transformation, SNV; normalization; Savitzky-Golay smoothing, SG; Savitzky-Golay 1 derivative, SG-1st; and Savitzky-Golay 2 derivative, SG-2nd), and three feature wavelength selection methods (competitive adaptive reweighted sampling, CARS; successive projections algorithm, SPA; and uninformative variable elimination, UVE) are compared. Partial least squares (PLS) and support vector machine (SVM) are used to construct prediction models for chilled fresh lamb meat quality. The results show that when rebound is used as a parameter, the SG-2nd-SPA-PLSR model has the highest accuracy, with a determination coefficient R²p of 0.94 for the prediction set. When TVB-N is used as a parameter, the MSC-UVE-SVM model has the highest accuracy, with an R²p of 0.95 for the prediction set. In conclusion, the use of near-infrared spectroscopic analysis enables rapid and non-destructive prediction and evaluation of lamb meat freshness, including its textural characteristics and TVB-N content under different modified atmosphere packaging. This study provides a theoretical basis and technical support for further encapsulating the models into portable devices and developing portable near-infrared spectrometers to rapidly determine lamb meat freshness.