Developing a Portable Fluorescence Imaging Device for Fish Freshness Detection.

Rapid detection of fish freshness is of vital importance to ensuring the safety of aquatic product consumption. Currently, the widely used optical detecting methods of fish freshness are faced with multiple challenges, including low detecting efficiency, high cost, large size and low integration of detecting equipment. This research aims to address these issues by developing a low-cost portable fluorescence imaging device for rapid fish freshness detection. The developed device employs ultraviolet-light-emitting diode (UV-LED) lamp beads (365 nm, 10 W) as excitation light sources, and a low-cost field programmable gate array (FPGA) board (model: ZYNQ XC7Z020) as the master control unit. The fluorescence images captured by a complementary metal oxide semiconductor (CMOS) camera are processed by the YOLOv4-Tiny model embedded in FPGA to obtain the ultimate results of fish freshness. The circuit for the YOLOv4-Tiny model is optimized to make full use of FPGA resources and to increase computing efficiency. The performance of the device is evaluated by using grass carp fillets as the research object. The average accuracy of freshness detection reaches up to 97.10%. Moreover, the detection time of below 1 s per sample and the overall power consumption of 47.1 W (including 42.4 W light source power consumption) indicate that the device has good real-time performance and low power consumption. The research provides a potential tool for fish freshness evaluation in a low-cost and rapid manner.

ADFSNet: An Adaptive Domain Feature Separation Network for the Classification of Wheat Seed Using Hyperspectral Images.

Wheat seed classification is a critical task for ensuring crop quality and yield. However, the characteristics of wheat seeds can vary due to variations in climate, soil, and other environmental factors across different years. Consequently, the present classification model is no longer adequate for accurately classifying novel samples. To tackle this issue, this paper proposes an adaptive domain feature separation (ADFS) network that utilizes hyperspectral imaging techniques for cross-year classification of wheat seed varieties. The primary objective is to improve the generalization ability of the model at a minimum cost. ADFS leverages deep learning techniques to acquire domain-irrelevant features from hyperspectral data, thus effectively addressing the issue of domain shifts across datasets. The feature spaces are divided into three parts using different modules. One shared module aligns feature distributions between the source and target datasets from different years, thereby enhancing the model's generalization and robustness. Additionally, two private modules extract class-specific features and domain-specific features. The transfer mechanism does not learn domain-specific features to reduce negative transfer and improve classification accuracy. Extensive experiments conducted on a two-year dataset comprising four wheat seed varieties demonstrate the effectiveness of ADFS in wheat seed classification. Compared with three typical transfer learning networks, ADFS can achieve the best accuracy of wheat seed classification with small batch samples updated, thereby addressing new seasonal variability.

Estimation of optical properties of turbid media using spatially resolved diffuse reflectance combined with LSTM-attention network.

The accurate estimation of the optical properties of turbid media by using a spatially resolved (SR) technique remains a challenging task due to measurement errors in the acquired spatially resolved diffuse reflectance (SRDR) and challenges in inversion model implementation. In this study, what we believe to be a novel data-driven model based on a long short-term memory network and attention mechanism (LSTM-attention network) combined with SRDR is proposed for the accurate estimation of the optical properties of turbid media. The proposed LSTM-attention network divides the SRDR profile into multiple consecutive and partially overlaps sub-intervals by using the sliding window technique, and uses the divided sub-intervals as the input of the LSTM modules. It then introduces an attention mechanism to evaluate the output of each module automatically and form a score coefficient, finally obtaining an accurate estimation of the optical properties. The proposed LSTM-attention network is trained with Monte Carlo (MC) simulation data to overcome the difficulty in preparing training (reference) samples with known optical properties. Experimental results of the MC simulation data showed that the mean relative error (MRE) with 5.59% for the absorption coefficient [with the mean absolute error (MAE) of 0.04 cm-1, coefficient of determination (R2) of 0.9982, and root mean square error (RMSE) of 0.058 cm-1] and 1.18% for the reduced scattering coefficient (with an MAE of 0.208 cm-1, R2 of 0.9996, and RMSE of 0.237 cm-1), which were significantly better than those of the three comparative models. The SRDR profiles of 36 liquid phantoms, collected using a hyperspectral imaging system that covered a wavelength range of 530-900 nm, were used to test the performance of the proposed model further. The results showed that the LSTM-attention model achieved the best performance (with the MRE of 14.89%, MAE of 0.022 cm-1, R2 of 0.9603, and RMSE of 0.026 cm-1 for the absorption coefficient; and the MRE of 9.76%, MAE of 0.732 cm-1, R2 of 0.9701, and RMSE of 1.470 cm-1for the reduced scattering coefficient). Therefore, SRDR combined with the LSTM-attention model provides an effective method for improving the estimation accuracy of the optical properties of turbid media.

Spatially Offset Raman Spectroscopy Combined with Attention-Based LSTM for Freshness Evaluation of Shrimp.

Optical detection of the freshness of intact in-shell shrimps is a well-known difficult task due to shell occlusion and its signal interference. The spatially offset Raman spectroscopy (SORS) is a workable technical solution for identifying and extracting subsurface shrimp meat information by collecting Raman scattering images at different distances from the offset laser incidence point. However, the SORS technology still suffers from physical information loss, difficulties in determining the optimum offset distance, and human operational errors. Thus, this paper presents a shrimp freshness detection method using spatially offset Raman spectroscopy combined with a targeted attention-based long short-term memory network (attention-based LSTM). The proposed attention-based LSTM model uses the LSTM module to extract physical and chemical composition information of tissue, weight the output of each module by an attention mechanism, and come together as a fully connected (FC) module for feature fusion and storage dates prediction. Modeling predictions by collecting Raman scattering images of 100 shrimps within 7 days. The R2, RMSE, and RPD of the attention-based LSTM model achieved 0.93, 0.48, and 4.06, respectively, which is superior to the conventional machine learning algorithm with manual selection of the optimal spatially offset distance. This method of automatically extracting information from SORS data by Attention-based LSTM eliminates human error and enables fast and non-destructive quality inspection of in-shell shrimp.

Evaluating performance of SORS-based subsurface signal separation methods using statistical replication Monte Carlo simulation.

Spatially offset Raman spectroscopy (SORS) is a depth-profiling technique with deep information enhancement. However, the interference of the surface layer cannot be eliminated without prior information. The signal separation method is an effective candidate for reconstructing pure subsurface Raman spectra, and there is still a lack of evaluation means for the signal separation method. Therefore, a method based on line-scan SORS combined with improved statistical replication Monte Carlo (SRMC) simulation was proposed to evaluate the effectiveness of food subsurface signal separation method. Firstly, SRMC simulates the photon flux in the sample, generates a corresponding number of Raman photons at each voxel of interest, and collects them by external map scanning. Then, 5625 groups of mixed signals with different optical characteristic parameters were convoluted with spectra of public database and application measurement and introduced into signal separation methods. The effectiveness and application range of the method were evaluated by the similarity between the separated signals and the source Raman spectra. Finally, the simulation results were verified by three packaged foods. FastICA method can effectively separate Raman signals from subsurface layer of food and thus promote deep quality evaluation of food.

A packaged food internal Raman signal separation method based on spatially offset Raman spectroscopy combined with FastICA.

Raman spectroscopy attempts to reflect food quality by characterizing molecular vibration and rotation. However, the blocking of optical signals by packaging materials and the interference of the optical signal generated by the packaging itself make the detection of internal food quality without destroying packaging highly difficult. In this regard, this paper proposes a novel packaged food internal signal separation based on spatially offset Raman spectroscopy (SORS) coupled with improved fast independent component analysis (FastICA). Firstly, the Raman scattering image of the packaged food with offset laser incident point was obtained. Then, the movable quadratic mean of information entropy was used to select the observation feature region of the image. Thirdly, the main independents decomposed by the optimized FastICA method were identified by spectral attenuation characteristics of the SORS peak signal. Finally, the non-negativity of the separated signal was ensured by baseline recognition and correction. The effectiveness of this method was verified by refactoring the similarity between the signal and the reference signal by testing three different packaging and four internal materials under standard experimental conditions. The applicability of the method was proved by the internal signal separation of three packaged foods on sale. The experimental results indicate that the proposed method can separate the Raman signal of packaged food and can be used as a pretreatment method and auxiliary analysis means for the detection of packaged food.

Performance Improvement of Handheld Raman Spectrometer for Mixture Components Identification Using Fuzzy Membership and Sparse Non-Negative Least Squares.

Due to the advantages of low price and convenience for end-users to conduct field-based, in-situ analysis, handheld Raman spectrometers are widely used in the identification of mixture components. However, the spectra collected by handheld Raman spectrometer usually have serious peak overlapping and spectral distortion, resulting in difficulties in component identification in the mixture. A novel method for mixture components identification based on the handheld Raman spectrometer was proposed in this study. The wavelet transform and Voight curve fitting method were used to extract the feature parameters from each Raman spectral peak, including Raman shift, maximum intensity, and full width at half-maximum (FWHM), and the similarities between the mixture and each substance in the database were calculated by fuzzy membership function based on extracted feature parameters. Then, the possible substances in the mixture were preliminarily screened out as candidates according to the similarity. Finally, the Raman spectra of these candidates were used to fit the spectra of the mixture, and the fitting coefficients obtained by sparse non-negative least squares algorithm were employed to further determine the suspected substance in the mixture. The Raman spectra of 190 liquid mixture samples and 158 powder mixture samples were collected using a handheld Raman spectrometer and these spectra were used to validate the identification performance of the proposed method. The proposed method could achieve good identification accuracy for different mixture samples. It shows that the proposed method is an effective way for the component identification in mixture by using a handheld Raman spectrometer.

Line-scan Raman scattering image and multivariate analysis for rapid and noninvasive detection of restructured beef.

The mean spectral (MS) features were extracted from Raman scattering images (RSI) of beef samples over the region of interest covering the spectral range of 789-1710cm-1 and the spatial offset range of 0-5 mm (for two sides of the incident laser). The RSI monitored the main change in the protein, amide bands, lipids, and amino acid residues. The classification model performance based on MS features compared the conventional Raman spectral features and confirmed the usefulness of RSI. Finally, the results showed that RSI technology is a reliable tool for rapid and noninvasive detection of restructured beef.

Detection of adulterated sugar with plastic packaging based on spatially offset Raman imaging.

BACKGROUND: The application of optical sensing technology in food adulteration detection has been extensively studied. However, due to the impact of packaging materials on the penetration depth of photons in foods and the interference from the optical properties of the packaging materials themselves, the use of optical sensing technology to detect packaged foods adulteration is still a well-known problem. RESULTS: The line-scan Raman imaging system was used to collect Raman hyperspectral images of adulterated sugars, made by mixing soft sugar and cheap glucose in seven different ratios. With the 0 and 3 mm (optimal offset distance) between line-laser source and scanning line, the Raman hyperspectral images of adulterated sugars covered by packaging plastic were acquired respectively. Using adulterated samples un-covered by packaging plastic as training samples, the Random Forest prediction model was developed, and excellent prediction performance was achieved for adulterated samples un-covered by packaging plastics. Compared with Raman data acquired with 0 mm offset distance, the performance of the prediction model was significantly improved, with 0.957 for coefficient of determination (R2 ), 0.413 for root mean square error of prediction (RMSEP), and 4.846 for residual predictive deviation (RPD), for adulterated samples with plastic packaging acquired with the 3 mm offset distance. CONCLUSIONS: The novel non-destructive method based on spatially offset Raman imaging technology, which can reduce the interference of packaging materials and enhance the signal of internal interesting materials, was proposed for detection of adulterated sugar with plastic packaging. The experiment results show that spatially offset imaging technology provides a candidate method for detecting adulteration of packaged foods. © 2021 Society of Chemical Industry.

M3GPSpectra: A novel approach integrating variable selection/construction and MLR modeling for quantitative spectral analysis.

Quantitative analysis of the physical or chemical properties of various materials by using spectral analysis technology combined with chemometrics has become an important method in the field of analytical chemistry. This method aims to build a model relationship (called prediction model) between feature variables acquired by spectral sensors and components to be measured. Feature selection or transformation should be conducted to reduce the interference of irrelevant information on the prediction model because original spectral feature variables contain redundant information and massive noise. Most existing feature selection and transformation methods are single linear or nonlinear operations, which easily lead to the loss of feature information and affect the accuracy of subsequent prediction models. This research proposes a novel spectroscopic technology-oriented, quantitative analysis model construction strategy named M3GPSpectra. This tool uses genetic programming algorithm to select and reconstruct the original feature variables, evaluates the performance of selected and reconstructed variables by using multivariate regression model (MLR), and obtains the best feature combination and the final parameters of MLR through iterative learning. M3GPSpectra integrates feature selection, linear/nonlinear feature transformation, and subsequent model construction into a unified framework and thus easily realizes end-to-end parameter learning to significantly improve the accuracy of the prediction model. When applied to six types of datasets, M3GPSpectra obtains 19 prediction models, which are compared with those obtained by seven linear or non-linear popular methods. Experimental results show that M3GPSpectra obtains the best performance among the eight methods tested. Further investigation verifies that the proposed method is not sensitive to the size of the training samples. Hence, M3GPSpectra is a promising spectral quantitative analytical tool.

Pose Estimation of Sweet Pepper through Symmetry Axis Detection.

The space pose of fruits is necessary for accurate detachment in automatic harvesting. This study presents a novel pose estimation method for sweet pepper detachment. In this method, the normal to the local plane at each point in the sweet-pepper point cloud was first calculated. The point cloud was separated by a number of candidate planes, and the scores of each plane were then separately calculated using the scoring strategy. The plane with the lowest score was selected as the symmetry plane of the point cloud. The symmetry axis could be finally calculated from the selected symmetry plane, and the pose of sweet pepper in the space was obtained using the symmetry axis. The performance of the proposed method was evaluated by simulated and sweet-pepper cloud dataset tests. In the simulated test, the average angle error between the calculated symmetry and real axes was approximately 6.5°. In the sweet-pepper cloud dataset test, the average error was approximately 7.4° when the peduncle was removed. When the peduncle of sweet pepper was complete, the average error was approximately 6.9°. These results suggested that the proposed method was suitable for pose estimation of sweet peppers and could be adjusted for use with other fruits and vegetables.

Expression of miR-9 in the serum of patients with acute ischemic stroke and its effect on neuronal damage.

BACKGROUND: This research was aimed to measure the expression of miR-9 in serum of acute ischemic stroke (AIS) patients and explore the role of miR-9 on OGD-induced neuronal damage. METHODS: In the present study, we measured the expression of miR-9 in serum of 65 AIS patients by real-time quantitative PCR (RT-qPCR) and the effect of miR-9 on oxygen-glucose deprivation (OGD)-induced neuronal injury was detected by CCK-8 in vitro. Western blot was used to measure the expression of protein. RESULTS: We found that the serum level of miR-9 in 65 AIS patients was significantly higher than that in control group (no-AIS), and was positively correlated with NIHSS score (r=0.627, P<0.001), infarct volume ((r=0.576, P<0.001), serum IL-8 (r=0.376, P=0.002), TNF-α (r=0.418, P<0.001), IL-6 (r=0.545, P<0.001), and IL-1ß (r=0.592, P<0.001). miR-9 expression levels were upregulated in cultured neurons with OGD treatment. The downregulation of miR-9 significantly alleviated OGD-induced neuronal injury. Dual-luciferase reporter assay demonstrated that SIRT1 was a target gene of miR-9, and miR-9 negatively regulated SIRT 1 expression and positively regulated p65 expression. CONCLUSIONS: All in all, our data showed that downregulation of miR-9 protected neurons against OGD/R-induced injury by the SIRT1-mediated NF-kB pathway.

Decomposition and correction overlapping peaks of LIBS using an error compensation method combined with curve fitting.

The laser induced breakdown spectroscopy (LIBS) technique is an effective method to detect material composition by obtaining the plasma emission spectrum. The overlapping peaks in the spectrum are a fundamental problem in the qualitative and quantitative analysis of LIBS. Based on a curve fitting method, this paper studies an error compensation method to achieve the decomposition and correction of overlapping peaks. The vital step is that the fitting residual is fed back to the overlapping peaks and performs multiple curve fitting processes to obtain a lower residual result. For the quantitative experiments of Cu, the Cu-Fe overlapping peaks in the range of 321-327 nm obtained from the LIBS spectrum of five different concentrations of CuSO4·5H2O solution were decomposed and corrected using curve fitting and error compensation methods. Compared with the curve fitting method, the error compensation reduced the fitting residual about 18.12-32.64% and improved the correlation about 0.86-1.82%. Then, the calibration curve between the intensity and concentration of the Cu was established. It can be seen that the error compensation method exhibits a higher linear correlation between the intensity and concentration of Cu, which can be applied to the decomposition and correction of overlapping peaks in the LIBS spectrum.

Application of Joint Skewness Algorithm to Select Optimal Wavelengths of Hyperspectral Image for Maize Seed Classification YANG Sai, ZHU Qi-bing*, HUANG Min

As an effective method for the nondestructive measurement of agricultural products quality, hyperspectral imaging technology has been widely studied in the field of seed classification and identification. Feature extraction and optimal wavelength selection are the two critical issues affecting the application of hyperspectral image in the field of seed identification. This study aimed to select optimal wavelengths from hyperspectral image data using joint skewness algorithm, so that they can be deployed in multispectral imaging-based inspection system for the automatic classification of maize seed. The hyperspectral images covering the wavelength range of 438～1 000 nm were acquired for 960 maize seeds including 10 varieties. After extracting the mean spectrum and entropy from the hyperspectral images, the joint skewness algorithm was used to select optimal wavelengths, and the classification models based on support vector machine were developed using the mean spectrum, entropy, and their combination, respectively. The experimental results indicated that the classification accuracy of the models developed by combination of the mean spectrum and entropy were higher than that of the mean spectrum or entropy for either full wavelengths or optimal wavelengths. The classification model for the combination of the mean spectrum and entropy based on the 10 optimal wavelengths selected by the joint skewness algorithm obtained 96.28% accuracy for test samples, with improvements of 4.30% and 20.38% over that of the mean spectrum and entropy, respectively, which was higher than the classification accuracy of the model that developed in the full wavelength (i.e., 93.47%). Meanwhile, the classification model based on joint skewness algorithm yielded the better classification accuracy than that of uninformative viable elimination algorithm, successive projections algorithm, and competitive adaptive reweighed sampling algorithm. This study made the online application of the hyperspectral image technology available for seed identification.

Penetration Depth Measurement of Near-Infrared Hyperspectral Imaging Light for Milk Powder.

The increasingly common application of the near-infrared (NIR) hyperspectral imaging technique to the analysis of food powders has led to the need for optical characterization of samples. This study was aimed at exploring the feasibility of quantifying penetration depth of NIR hyperspectral imaging light for milk powder. Hyperspectral NIR reflectance images were collected for eight different milk powder products that included five brands of non-fat milk powder and three brands of whole milk powder. For each milk powder, five different powder depths ranging from 1 mm-5 mm were prepared on the top of a base layer of melamine, to test spectral-based detection of the melamine through the milk. A relationship was established between the NIR reflectance spectra (937.5-1653.7 nm) and the penetration depth was investigated by means of the partial least squares-discriminant analysis (PLS-DA) technique to classify pixels as being milk-only or a mixture of milk and melamine. With increasing milk depth, classification model accuracy was gradually decreased. The results from the 1-mm, 2-mm and 3-mm models showed that the average classification accuracy of the validation set for milk-melamine samples was reduced from 99.86% down to 94.93% as the milk depth increased from 1 mm-3 mm. As the milk depth increased to 4 mm and 5 mm, model performance deteriorated further to accuracies as low as 81.83% and 58.26%, respectively. The results suggest that a 2-mm sample depth is recommended for the screening/evaluation of milk powders using an online NIR hyperspectral imaging system similar to that used in this study.

[Purity Detection Model Update of Maize Seeds Based on Active Learning].

Seed purity reflects the degree of seed varieties in typical consistent characteristics, so it is great important to improve the reliability and accuracy of seed purity detection to guarantee the quality of seeds. Hyperspectral imaging can reflect the internal and external characteristics of seeds at the same time, which has been widely used in nondestructive detection of agricultural products. The essence of nondestructive detection of agricultural products using hyperspectral imaging technique is to establish the mathematical model between the spectral information and the quality of agricultural products. Since the spectral information is easily affected by the sample growth environment, the stability and generalization of model would weaken when the test samples harvested from different origin and year. Active learning algorithm was investigated to add representative samples to expand the sample space for the original model, so as to implement the rapid update of the model's ability. Random selection (RS) and Kennard-Stone algorithm (KS) were performed to compare the model update effect with active learning algorithm. The experimental results indicated that in the division of different proportion of sample set (1:1, 3:1, 4:1), the updated purity detection model for maize seeds from 2010 year which was added 40 samples selected by active learning algorithm from 2011 year increased the prediction accuracy for 2011 new samples from 47%, 33.75%, 49% to 98.89%, 98.33%, 98.33%. For the updated purity detection model of 2011 year, its prediction accuracy for 2010 new samples increased by 50.83%, 54.58%, 53.75% to 94.57%, 94.02%, 94.57% after adding 56 new samples from 2010 year. Meanwhile the effect of model updated by active learning algorithm was better than that of RS and KS. Therefore, the update for purity detection model of maize seeds is feasible by active learning algorithm.

Determination of sucrose content in sugar beet by portable visible and near-infrared spectroscopy.

Visible and near-infrared spectra in interactance mode were acquired for intact and sliced beet samples, using two portable spectrometers for the spectral regions of 400-1100 nm and 900-1600 nm, respectively. Sucrose prediction models for intact and sliced beets were developed and then validated. The spectrometer for 400-1100 nm was able to predict the sucrose content with correlations of prediction (rp) of 0.80 and 0.88 and standard errors of prediction (SEPs) of 0.89% and 0.70%, for intact beets and beet slices, respectively. The spectrometer for 900-1600 nm had rp values of 0.74 and 0.88 and SEPs of 1.02% and 0.69% for intact beets and beet slices. These results showed the feasibility of using the portable spectrometer to predict the sucrose content of beet slices. Using simple correlation analysis, the study also identified important wavelengths that had strong correlation with the sucrose content.

[Maize seed identification using hyperspectral imaging and SVDD algorithm].

The sufficiency of feature extraction and the rationality of classifier design are two key issues affecting the accuracy of maize seed recognition. In the present study, the hyperspectral images of maize seeds were acquired using hyperspectral image system, and the image entropy of maize seeds for each wavelength was extracted as classification features. Then, support vector data description (SVDD) algorithm was used to develop the classifier model for each variety of maize seeds. The SVDD models yielded 94.14% average test accuracy for known variety samples and 92.28% average test accuracy for new variety samples, respectively. The simulation results showed that the proposed method implemented accurate identification of maize seeds and solved the problem of misclassification by the traditional classification algorithm for new variety maize seeds.

[Feature extraction of hyperspectral scattering image for apple mealiness based on singular value decomposition].

Apple mealiness is an important sensory parameter for classification of apple quality. Hyperspectral scattering technique was investigated for noninvasive detection of apple mealiness. A singular value decomposition (SVD) method was proposed to extract the feature/ or singular values of the hyperspectral scattering images between 600 and 1000 nm for 20 mm distance including 81 wavelengths. As characteristic parameters of apple mealiness, singular values were applied to develop the classification model coupled with partial least squares discriminant analysis (PLSDA) using the samples from different origin and different storage conditions. The classification accuracies for the two-class ("mealy" and "non-mealy") model were between 76.1% and 80.6% better than mean method (75.3%-76.5%). The results indicated that SVD method was potentially useful for the feature extraction of hyperspectral scattering images and the model developed with these features can detect the mealy and non-mealy apple, but the classification accuracies need to be improved.

[LLE-SVM classification of apple mealiness based on hyperspectral scattering image].

Apple mealiness degree is an important factor for its internal quality. hyperspectral scattering, as a promising technique, was investigated for noninvasive measurement of apple mealiness. In the present paper, a locally linear embedding (LLE) coupled with support vector machine (SVM) was proposed to achieve classification because of large number of image data. LLE is a nonlinear lowering dimension method, which reveals the structure of the global nonlinearity by the local linear joint. This method can effectively calculate high-dimensional input data embedded in a low-dimensional space manifold. The dimension reduction of hyperspectral data was classified by SVM. Comparing the LLE-SVM classification method with the traditional SVM classification, the results indicated that the training accuracy obtained with the LLE-SVM was higher than that just with SVM; and the testing accuracy of the classifier changed a little before and after dimensionality reduction, and the range of fluctuation was less than 5%. It is expected that LLE-SVM method would provide an effective classification method for apple mealiness nondestructive detection using hyperspectral scattering image technique.