Identification of geographical origins of Radix Paeoniae Alba using hyperspectral imaging with deep learning-based fusion approaches

Simultaneously predicting SPAD and water content in rice leaves using hyperspectral imaging with deep multi-task regression and transfer component analysis

BACKGROUND

Water content and chlorophyll content are important indicators for monitoring rice growth status. Simultaneous detection of water content and chlorophyll content is of significance. Different varieties of rice show differences in phenotype, resulting in the difficulties of establishing a universal model. In this study, hyperspectral imaging was used to detect the Soil and Plant Analyzer Development (SPAD) values and water content of fresh rice leaves of three rice varieties (Jiahua 1, Xiushui 121 and Xiushui 134).

RESULTS

Both partial least squares regression and convolutional neural networks were used to establish single-task and multi-task models. Transfer component analysis (TCA) was used as transfer learning to learn the common features to achieve an approximate identical distribution between any two varieties. Single-task and multi-task models were also built using the features of the source domain, and these models were applied to the target domain. These results indicated that for models of each rice variety the prediction accuracy of most multi-task models was close to that of single-task models. As for TCA, the results showed that the single-task model achieved good performance for all transfer learning tasks.

CONCLUSION

Compared with the original model, good and differentiated results were obtained for the models using features learned by TCA for both the source domain and target domain. The multi-task models could be constructed to predict SPAD values and water content simultaneously and then transferred to another rice variety, which could improve the efficiency of model construction and realize rapid detection of rice growth indicators. © 2024 Society of Chemical Industry.

Fusion of convolutional neural network with XGBoost feature extraction for predicting multi-constituents in corn using near infrared spectroscopy

Near-infrared (NIR) spectroscopy has been widely utilized to predict multi-constituents of corn in agriculture. However, directly extracting constituent information from the NIR spectra is challenging due to many issues such as broad absorption band, overlapping and non-specific nature. To solve these problems and extract implicit features from the raw data of NIR spectra to improve performance of quantitative models, a one-dimensional shallow convolutional neural network (CNN) model based on an eXtreme Gradient Boosting (XGBoost) feature extraction method was proposed in this paper. The leaf node feature information in the XGBoost was encoded and reconstructed to obtain the implicit features of raw data in the NIR spectra. A two-parametric Swish (TSwish or TS) activation function was proposed to improve the performance of CNN, and the elastic net (EN) was also applied to avoid the overfitting problem of the CNN model. Performance of the developed XGBoost-CNN-TS-EN model was evaluated using two public NIR spectroscopy datasets of corn and soil, and the obtained determination coefficients (R2) for moisture, oil, protein, and starch of the corn on test set were 0.993, 0.991, 0.998, and 0.992, respectively, with that of the soil organic matter being 0.992. The XGBoost-CNN-TS-EN model exhibits superior stability, good prediction accuracy, and generalization ability, demonstrating its great potentials for quantitative analysis of multi-constituents in spectroscopic applications.

Recent advances of machine learning in the geographical origin traceability of food and agro-products: A review

The geographical origin traceability of food and agro-products has been attracted worldwide. Especially with the rise of machine learning (ML) technology, it provides cutting-edge solutions to erstwhile intractable issues to identify the origin of food and agro-products. By utilizing advanced algorithms, ML can extract feature information of food and agro-products that is closely related to origin and, more accurately, identify and trace their origins, which is of great significance to the entire food and agriculture industry. This paper provides a comprehensive overview of the state-of-the-art applications of ML in the geographical origin traceability of food and agro-products. First, commonly used ML methods are summarized. The paper then outlines the whole process of preparation for modeling, model training as well as model evaluation for building traceability models-based ML. Finally, recent applications of ML combined with different traceability techniques in the field of food and agro-products are revisited. Although ML has made many achievements in solving the geographical origin traceability problem of food and agro-products, it still has great development potential. For example, the application of ML is yet insufficient in the geographical origin traceability using DNA or computer vision techniques. The ability of ML to predict the geographical origin of food and agro-products can be further improved, for example, by increasing model interpretability, incorporating data fusion strategies, and others.

The combination of HSI and NMR techniques with deep learning for identification of geographical origin and GI markers of Lycium barbarum L

Lycium barbarum L. (L. barbarum) is renowned worldwide for its nutritional and medicinal benefits. Rapid and accurate identification of L.barbarum's geographic origin is essential because its nutritional content, medicinal efficacy, and market price significantly vary by region. This study proposes an innovative method combining hyperspectral imaging (HSI), nuclear magnetic resonance (NMR), and an improved ResNet-34 deep learning model to accurately identify the geographical origin and geographical indication (GI) markers of L.barbarum. The deep learning model achieved a 95.63% accuracy, surpassed traditional methods by 6.26% and reduced runtime by 29.9% through SHapley Additive exPlanations (SHAP)-based feature selection. Pearson correlation analysis between GI markers and HSI characteristic wavelengths enhanced the interpretability of HSI data and further reduced runtime by 33.99%. This work lays the foundation for portable multispectral devices, offering a rapid, accurate, and cost-effective solution for quality assurance and market regulation of L.barbarum products.

Combining Vis-NIR and NIR hyperspectral imaging techniques with a data fusion strategy for prediction of norfloxacin residues in mutton

Norfloxacin is an antibacterial compound that belongs to the fluoroquinolone family. Currently, hyperspectral imaging (HSI) for the detection of antibiotic residues focuses mostly on individual systems. Attempts to integrate different HSI systems with complementary spectral ranges are still lacking. This study investigates the feasibility of applying data fusion strategies with two HSI techniques (Visible near-infrared and near-infrared) in combination to predict norfloxacin residue levels in mutton. Spectral data from the two spectral techniques were analyzed using partial least squares regression (PLSR), support vector regression (SVR) and stochastic configuration networks (SCN), respectively, and the two data fusion strategies were fused at the data level (low-level fusion) and feature level (middle-level fusion, mid-level fusion). The results indicated that the modeling performance of the two fused datasets was better than that of the individual systems. Mid-level fusion data achieved the best model based on uninformative variable elimination (UVE) combined with SCN, in which the determination coefficient of prediction set (R2p) of 0.9312, (root mean square error of prediction set) RMSEP of 0.3316 and residual prediction deviation (RPD) of 2.7434, in comparison with all others. Therefore, two HSI systems with complementary spectral ranges, combined with data fusion strategies and feature selection, could be used synergistically to improve the detection of norfloxacin residues. This study may provide a valuable reference for the non-destructive detection of antibiotic residues in meat.

Comprehensive characterization of anthraquinones in Damnacanthus indicus using mass spectrometry molecular networking and metabolomics-based herb discrimination

Damnacanthus indicus is a widely used folk medicine in China, renowned for its various bioactivities. The key active components, anthraquinones, have not been comprehensively profiled due to their complex chemical nature. Establishing a high-throughput strategy to systematically characterize these anthraquinones is essential. Additionally, the cultivation of D. indicus across various provinces results in significant quality differences in the harvested herbs. Thus, developing an effective strategy to distinguish herbs from different regions and identify characteristic chemical markers for quality evaluation and control is crucial. In this study, a strategy based on ultra-high performance liquid chromatography-mass spectrometry (UHPLC-MS) was employed to systematically characterize the chemical composition of D. indicus. Mass spectrometry molecular networking was utilized to rapidly recognize and identify anthraquinones. Principal component analysis (PCA) was applied to cluster the herbs from different habitats, while partial least square discriminant analysis (PLS-DA) was used to screen for chemical markers distinguishing herb origins. The result showed that a total of 112 anthraquinones and 66 non-anthraquinone compounds were identified in D. indicus. The biosynthetic pathways of anthraquinones in this herb were proposed. PCA grouped 15 batches of herbs from different origins into three clusters, corresponding to the climate types of their habitats. PLS-DA identified 27 significant chemical markers that could robustly distinguish the geographical origins of the herbs. This study provides a valuable reference for the quality evaluation and control of D. indicus and offers a scientific basis for the pharmacological research and rational utilization of these medicinal resources.

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.

Accurate and visualiable discrimination of Chenpi age using 2D-CNN and Grad-CAM++ based on infrared spectral images

Dried tangerine peel ("Chenpi"), has numerous clinical and nutritional benefits, with its quality being significantly influenced by its storage age, referred to as "Chen Jiu Zhe Liang" in Chinese. Concequently, the rapid and accurate identification of Chenpi's age is important for consumers. In this study, Fourier transform infrared spectroscopy (FTIR) was employed to capture spectral images of Chenpi. These FTIR images were then analyzed using a two-dimensional convolutional neural networks (2D-CNN) model, achieving a discrimination accuracy of 97.92%. To address the "black box" nature of the 2D-CNN, Gradient-weighted Class Activation Mapping Plus Plus (Grad-CAM++) was utilized to highlight the important regions contributing to the model's performance. Additionally, six other machine learning models were developped using features identified by the 2D-CNN to validate their effectiveness. The results demonstrated that the combination of FTIR spectral images and 2D-CNN provides a highly effective method for accurately determining the age of Chenpi.

Geographical discrimination of Asian red pepper powders using 1H NMR spectroscopy and deep learning-based convolution neural networks

This study investigated an innovative approach to discriminate the geographical origins of Asian red pepper powders by analyzing one-dimensional 1H NMR spectra through a deep learning-based convolution neural network (CNN). 1H NMR spectra were collected from 300 samples originating from China, Korea, and Vietnam and used as input data. Principal component analysis - linear discriminant analysis and support vector machine models were employed for comparison. Bayesian optimization was used for hyperparameter optimization, and cross-validation was performed to prevent overfitting. As a result, all three models discriminated the origins of the test samples with over 95 % accuracy. Specifically, the CNN models achieved a 100 % accuracy rate. Gradient-weighted class activation mapping analysis verified that the CNN models recognized the origins of the samples based on variations in metabolite distributions. This research demonstrated the potential of deep learning-based classification of 1H NMR spectra as an accurate and reliable approach for determining the geographical origins of various foods.

Identification and Classification of Coix seed Storage Years Based on Hyperspectral Imaging Technology Combined with Deep Learning

Developing a fast and non-destructive methodology to identify the storage years of Coix seed is important in safeguarding consumer well-being. This study employed the utilization of hyperspectral imaging (HSI) in conjunction with conventional machine learning techniques such as support vector machines (SVM), k-nearest neighbors (KNN), random forest (RF), extreme gradient boosting (XGBoost), as well as the deep learning method of residual neural network (ResNet), to establish identification models for Coix seed samples from different storage years. Under the fusion-based modeling approach, the model's classification accuracy surpasses that of visible to near infrared (VNIR) and short-wave infrared (SWIR) spectral modeling individually. The classification accuracy of the ResNet model and SVM exceeds that of other conventional machine learning models (KNN, RF, and XGBoost). Redundant variables were further diminished through competitive adaptive reweighted sampling feature wavelength screening, which had less impact on the model's accuracy. Upon validating the model's performance using an external validation set, the ResNet model yielded more satisfactory outcomes, exhibiting recognition accuracy exceeding 85%. In conclusion, the comprehensive results demonstrate that the integration of deep learning with HSI techniques effectively distinguishes Coix seed samples from different storage years.

Non-destructive prediction of isoflavone and starch by hyperspectral imaging and deep learning in Puerariae Thomsonii Radix

Accurate assessment of isoflavone and starch content in Puerariae Thomsonii Radix (PTR) is crucial for ensuring its quality. However, conventional measurement methods often suffer from time-consuming and labor-intensive procedures. In this study, we propose an innovative and efficient approach that harnesses hyperspectral imaging (HSI) technology and deep learning (DL) to predict the content of isoflavones (puerarin, puerarin apioside, daidzin, daidzein) and starch in PTR. Specifically, we develop a one-dimensional convolutional neural network (1DCNN) model and compare its predictive performance with traditional methods, including partial least squares regression (PLSR), support vector regression (SVR), and CatBoost. To optimize the prediction process, we employ various spectral preprocessing techniques and wavelength selection algorithms. Experimental results unequivocally demonstrate the superior performance of the DL model, achieving exceptional performance with mean coefficient of determination (R2) values surpassing 0.9 for all components. This research underscores the potential of integrating HSI technology with DL methods, thereby establishing the feasibility of HSI as an efficient and non-destructive tool for predicting the content of isoflavones and starch in PTR. Moreover, this methodology holds great promise for enhancing efficiency in quality control within the food industry.

Rice seed vigor detection based on near-infrared hyperspectral imaging and deep transfer learning

Vigor is one of the important factors that affects rice yield and quality. Rapid and accurate detection of rice seed vigor is of great importance for rice production. In this study, near-infrared hyperspectral imaging technique and transfer learning were combined to detect rice seed vigor. Four varieties of artificial-aged rice seeds (Yongyou12, Yongyou1540, Suxiangjing100, and Longjingyou1212) were studied. Different convolutional neural network (CNN) models were built to detect the vigor of the rice seeds. Two transfer strategies, fine-tuning and MixStyle, were used to transfer knowledge among different rice varieties for vigor detection. The experimental results showed that the convolutional neural network model of Yongyou12 classified the vigor of Yongyou1540, Suxiangjing100, and Longjingyou1212 through MixStyle transfer knowledge, and the accuracy reached 90.00%, 80.33%, and 85.00% in validation sets, respectively, which was better or close to the initial modeling performances of each variety. MixStyle statistics are based on probabilistic mixed instance-level features of cross-source domain training samples. When training instances, new domains can be synthesized, which increases the domain diversity of the source domain, thereby improving the generalization ability of the trained model. This study would help rapid and accurate detection of a large varieties of crop seeds.

Rapid Prediction of Adulteration Content in Atractylodis rhizoma Based on Data and Image Features Fusions from Near-Infrared Spectroscopy and Hyperspectral Imaging Techniques

Atractylodis rhizoma (AR) is an herb and food source with great economic, medicinal, and ecological value. Atractylodes chinensis (DC.) Koidz. (AC) and Atractylodes lancea (Thunb.) DC. (AL) are its two botanical sources. The commercial fraud of AR adulterated with Atractylodes japonica Koidz. ex Kitam (AJ) frequently occurs in pursuit of higher profit. To quickly determine the content of adulteration in AC and AL powder, two spectroscopic techniques, near-infrared spectroscopy (NIRS) and hyperspectral imaging (HSI), were introduced. The partial least squares regression (PLSR) algorithm was selected for predictive modeling of AR adulteration levels. Preprocessing and feature variable extraction were used to optimize the prediction model. Then data and image feature fusions were developed to obtain the best predictive model. The results showed that if only single-spectral techniques were considered, NIRS was more suitable for both tasks than HSI techniques. In addition, by comparing the models built after the data fusion of NIRS and HSI with those built by the single spectrum, we found that the mid-level fusion strategy obtained the best models in both tasks. On this basis, combined with the color-texture features, the prediction ability of the model was further optimized. Among them, for the adulteration level prediction task of AC, the best strategy was combining MLF data (at CARS level) and color-texture features (C-TF), at which time the R2T, RMSET, R2P, and RMSEP were 99.85%, 1.25%, 98.61%, and 5.06%, respectively. For AL, the best approach was combining MLF data (at SPA level) and C-TF, with the highest R2T (99.92%) and R2P (99.00%), as well as the lowest RMSET (1.16%) and RMSEP (2.16%). Therefore, combining data and image features from NIRS and HSI is a potential strategy to predict the adulteration content quickly, non-destructively, and accurately.