Application of hyperspectral imaging and chemometrics for variety classification of maize seeds

An improved deep convolutional generative adversarial network for quantification of catechins in fermented black tea

The rapid and non-destructive quantification of catechins in fermented black tea is crucial for evaluating the quality of black tea. The combination of hyperspectral imaging and chemometrics has been applied for quantitative detection, but its performance is usually constrained by the limited dataset size. Targeted at the challenge of insufficient samples in regression analysis of catechins, this study proposes an improved deep convolutional generative adversarial network with labeling module, named as DCGAN-L for hyperspectral data augmentation. The DCGAN-L consists of the spectral and label generating modules. First the synthetic spectra were generated, and an indicator was proposed to evaluate their quality. Then, the corresponding label values were generated, including epicatechin gallate (ECG), epicatechin (EGC), catechin (C), and total catechin (CC). For label generating, the Euclidean distances between the synthetic spectrum and all true spectra were measured, followed by allocating weights for calculating the label values based on these distances. Subsequently, the training dataset was augmented with the generated synthetic data. The effect of data augmentation was finally evaluated based on two regression models of random forest (RF) and broad learning system (BLS) for the quantification of catechins. Compared with the results before data augmentation, the average R2 of RF and BLS models increased by 0.044 and 0.164, respectively. The proposed DCGAN-L model allows for the rapid, non-destructive quantification of catechins in black tea in the case of limited sample size.

Grain Protein Content Phenotyping in Rice via Hyperspectral Imaging Technology and a Genome-Wide Association Study

Efficient and accurate acquisition of the rice grain protein content (GPC) is important for selecting high-quality rice varieties, and remote sensing technology is an attractive potential method for this task. However, the majority of multispectral sensors are poor predictors of GPC due to their broad spectral bands. Hyperspectral technology provides a new analytical technology for bridging the gap between phenomics and genomics. However, the small size of typical datasets is a constraint for model construction for estimating GPC, limiting their accuracy and reducing their ability to generalize to a wide range of varieties. In this study, we used hyperspectral data of rice grains from 515 japonica varieties and deep convolution generative adversarial networks (DCGANs) to generate simulated data to improve the model accuracy. Features sensitive to GPC were extracted after applying a continuous wavelet transform (CWT), and the estimated GPC model was constructed by partial least squares regression (PLSR). Finally, a genome-wide association study (GWAS) was applied to the measured and generated datasets to detect GPC loci. The results demonstrated that the simulated GPC values generated after 8,000 epochs were closest to the measured values. The wavelet feature (WF1743, 2), obtained from the data with the addition of 200 simulated samples, exhibited the highest GPC estimation accuracy (R 2 = 0.58 and RRMSE = 6.70%). The GWAS analysis showed that the estimated values based on the simulated data detected the same loci as the measured values, including the OsmtSSB1L gene related to grain storage protein. This study provides a new technique for the efficient genetic study of phenotypic traits in rice based on hyperspectral technology.

Fully connected-convolutional (FC-CNN) neural network based on hyperspectral images for rapid identification of P. ginseng growth years

P. ginseng is a precious traditional Chinese functional food, which is used for both medicinal and food purposes, and has various effects such as immunomodulation, anti-tumor and anti-oxidation. The growth year of P. ginseng has an important impact on its medicinal and economic values. Fast and nondestructive identification of the growth year of P. ginseng is crucial for its quality evaluation. In this paper, we propose a FC-CNN network that incorporates spectral and spatial features of hyperspectral images to characterize P. ginseng from different growth years. The importance ranking of the spectra was obtained using the random forest method for optimal band selection. Based on the hyperspectral reflectance data of P. ginseng after radiometric calibration and the images of the best five VNIR bands and five SWIR bands selected, the year-by-year identification of P. ginseng age and its identification experiments for food and medicinal purposes were conducted, and the FC-CNN network and its FCNN and CNN branch networks were tested and compared in terms of their effectiveness in the identification of P. ginseng growth years. It has been experimentally verified that the best year-by-year recognition was achieved by utilizing images from five visible and near-infrared important bands and all spectral curves, and the recognition accuracy of food and medicinal use reached 100%. The FC-CNN network is significantly better than its branching model in the effect of edible and medicinal identification. The results show that for P. ginseng growth year identification, VNIR images have much more useful information than SWIR images. Meanwhile, the FC-CNN network utilizing the spectral and spatial features of hyperspectral images is an effective method for the identification of P. ginseng growth year.

Maize seeds forecasting with hybrid directional and bi-directional long short-term memory models

The purity of the seeds is one of the important factors that increase the yield. For this reason, the classification of maize cultivars constitutes a significant problem. Within the scope of this study, six different classification models were designed to solve this problem. A special dataset was created to be used in the models designed for the study. The dataset contains a total of 14,469 images in four classes. Images belong to four different maize types, BT6470, CALIPOS, ES_ARMANDI, and HIVA, taken from the BIOTEK company. AlexNet and ResNet50 architectures, with the transfer learning method, were used in the models created for the image classification. In order to improve the classification success, LSTM (Directional Long Short-Term Memory) and BiLSTM (Bi-directional Long Short-Term Memory) algorithms and AlexNet and ResNet50 architectures were hybridized. As a result of the classifications, the highest classification success was obtained from the ResNet50+BiLSTM model with 98.10%.

Non-destructive identification of Pseudostellaria heterophylla from different geographical origins by Vis/NIR and SWIR hyperspectral imaging techniques

The composition of Pseudostellaria heterophylla (Tai-Zi-Shen, TZS) is greatly influenced by the growing area of the plants, making it significant to distinguish the origins of TZS. However, traditional methods for TZS origin identification are time-consuming, laborious, and destructive. To address this, two or three TZS accessions were selected from four different regions of China, with each of these resources including distinct quality grades of TZS samples. The visible near-infrared (Vis/NIR) and short-wave infrared (SWIR) hyperspectral information from these samples were then collected. Fast and high-precision methods to identify the origins of TZS were developed by combining various preprocessing algorithms, feature band extraction algorithms (CARS and SPA), traditional two-stage machine learning classifiers (PLS-DA, SVM, and RF), and an end-to-end deep learning classifier (DCNN). Specifically, SWIR hyperspectral information outperformed Vis/NIR hyperspectral information in detecting geographic origins of TZS. The SPA algorithm proved particularly effective in extracting SWIR information that was highly correlated with the origins of TZS. The corresponding FD-SPA-SVM model reduced the number of bands by 77.2% and improved the model accuracy from 97.6% to 98.1% compared to the full-band FD-SVM model. Overall, two sets of fast and high-precision models, SWIR-FD-SPA-SVM and SWIR-FD-DCNN, were established, achieving accuracies of 98.1% and 98.7% respectively. This work provides a potentially efficient alternative for rapidly detecting the origins of TZS during actual production.

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.

Current Application of Advancing Spectroscopy Techniques in Food Analysis: Data Handling with Chemometric Approaches

In today's era of increased food consumption, consumers have become more demanding in terms of safety and the quality of products they consume. As a result, food authorities are closely monitoring the food industry to ensure that products meet the required standards of quality. The analysis of food properties encompasses various aspects, including chemical and physical descriptions, sensory assessments, authenticity, traceability, processing, crop production, storage conditions, and microbial and contaminant levels. Traditionally, the analysis of food properties has relied on conventional analytical techniques. However, these methods often involve destructive processes, which are laborious, time-consuming, expensive, and environmentally harmful. In contrast, advanced spectroscopic techniques offer a promising alternative. Spectroscopic methods such as hyperspectral and multispectral imaging, NMR, Raman, IR, UV, visible, fluorescence, and X-ray-based methods provide rapid, non-destructive, cost-effective, and environmentally friendly means of food analysis. Nevertheless, interpreting spectroscopy data, whether in the form of signals (fingerprints) or images, can be complex without the assistance of statistical and innovative chemometric approaches. These approaches involve various steps such as pre-processing, exploratory analysis, variable selection, regression, classification, and data integration. They are essential for extracting relevant information and effectively handling the complexity of spectroscopic data. This review aims to address, discuss, and examine recent studies on advanced spectroscopic techniques and chemometric tools in the context of food product applications and analysis trends. Furthermore, it focuses on the practical aspects of spectral data handling, model construction, data interpretation, and the general utilization of statistical and chemometric methods for both qualitative and quantitative analysis. By exploring the advancements in spectroscopic techniques and their integration with chemometric tools, this review provides valuable insights into the potential applications and future directions of these analytical approaches in the food industry. It emphasizes the importance of efficient data handling, model development, and practical implementation of statistical and chemometric methods in the field of food analysis.

Y-Net: Identification of Typical Diseases of Corn Leaves Using a 3D-2D Hybrid CNN Model Combined with a Hyperspectral Image Band Selection Module

Corn diseases are one of the significant constraints to high-quality corn production, and accurate identification of corn diseases is of great importance for precise disease control. Corn anthracnose and brown spot are typical diseases of corn, and the early symptoms of the two diseases are similar, which can be easily misidentified by the naked eye. In this paper, to address the above problems, a three-dimensional-two-dimensional (3D-2D) hybrid convolutional neural network (CNN) model combining a band selection module is proposed based on hyperspectral image data, which combines band selection, attention mechanism, spatial-spectral feature extraction, and classification into a unified optimization process. The model first inputs hyperspectral images to both the band selection module and the attention mechanism module and then sums the outputs of the two modules as inputs to a 3D-2D hybrid CNN, resulting in a Y-shaped architecture named Y-Net. The results show that the spectral bands selected by the band selection module of Y-Net achieve more reliable classification performance than traditional feature selection methods. Y-Net obtained the best classification accuracy compared to support vector machines, one-dimensional (1D) CNNs, and two-dimensional (2D) CNNs. After the network pruned the trained Y-Net, the model size was reduced to one-third of the original size, and the accuracy rate reached 98.34%. The study results can provide new ideas and references for disease identification of corn and other crops.

Deep Learning Combined with Hyperspectral Imaging Technology for Variety Discrimination of Fritillaria thunbergii

Traditional Chinese herbal medicine (TCHM) plays an essential role in the international pharmaceutical industry due to its rich resources and unique curative properties. The flowers, stems, and leaves of Fritillaria contain a wide range of phytochemical compounds, including flavonoids, essential oils, saponins, and alkaloids, which may be useful for medicinal purposes. Fritillaria thunbergii Miq. Bulbs are commonly used in traditional Chinese medicine as expectorants and antitussives. In this paper, a feasibility study is presented that examines the use of hyperspectral imaging integrated with convolutional neural networks (CNN) to distinguish twelve (12) Fritillaria varieties (n = 360). The performance of support vector machines (SVM) and partial least squares-discriminant analysis (PLS-DA) was compared with that of convolutional neural network (CNN). Principal component analysis (PCA) was used to assess the presence of cluster trends in the spectral data. To optimize the performance of the models, cross-validation was used. Among all the discriminant models, CNN was the most accurate with 98.88%, 88.89% in training and test sets, followed by PLS-DA and SVM with 92.59%, 81.94% and 99.65%, 79.17%, respectively. The results obtained in the present study revealed that application of HSI in conjunction with the deep learning technique can be used for classification of Fritillaria thunbergii varieties rapidly and non-destructively.

Computer-Aided Multiclass Classification of Corn from Corn Images Integrating Deep Feature Extraction

Corn has great importance in terms of production in the field of agriculture and animal feed. Obtaining pure corn seeds in corn production is quite significant for seed quality. For this reason, the distinction of corn seeds that have numerous varieties plays an essential role in marketing. This study was conducted with 14,469 images of BT6470, Calipso, Es_Armandi, and Hiva types of corn licensed by BIOTEK. The classification of images was carried out in three stages. At the first stage, deep feature extraction of the four types of corn images was performed with the pretrained CNN model SqueezeNet 1000 deep features were obtained for each image. In the second stage, in order to reduce these features obtained from deep feature extraction with SqueezeNet, separate feature selection processes were performed with the Bat Optimization (BA), Whale Optimization (WOA), and Gray Wolf Optimization (GWO) algorithms among optimization algorithms. Finally, in the last stage, the features obtained from the first and second stages were classified by using the machine learning methods Decision Tree (DT), Naive Bayes (NB), multi-class Support Vector Machine (mSVM), k-Nearest Neighbor (KNN), and Neural Network (NN). In the classification processes of the features obtained in the first stage, the mSVM model has achieved the highest classification success with 89.40%. In the second stage, as a result of the classifications performed through the active features selected by using three types of feature selection algorithms (BA, WOA, GWO), the classification success obtained with the mSVM model was 88.82%, 88.72%, and 88.95%, respectively. The classification accuracies of the tested methods and the classification accuracies obtained in the first stage are close to each other in terms of classification success. However, with the algorithms used in feature selection, successful classification processes have been carried out with fewer features and in a shorter time. The results of the study, in which classification was carried out in the inexpensive, the objective, and the shorter time of processing for the corn types, present a different perspective in terms of classification performance.

Prediction of oil content in single maize kernel based on hyperspectral imaging and attention convolution neural network

An attention (A) based convolutional neural network regression (CNNR) model, namely ACNNR, was proposed to combine hyperspectral imaging to predict oil content in single maize kernel. During the period, a reflectance HSI system was used to collect hyperspectral images of embryo side and non-embryo side of single maize kernel, and the performances of CNNR (without attention mechanism), ACNNR and partial least squares regression (PLSR) were compared. For PLSR, a series of spectral preprocessing and dimensionality reduction methods were used to finally determine the optimal hybrid PLSR model. Whereas for CNNR and ACNNR, only raw spectra were used as their inputs. The results showed that embryo side was more suitable for developing regression models; the attentional mechanism was helpful to reduce the error of prediction, making ACNNR performed best (coefficient of determination of prediction = 0.9198). Overall, the proposed method did not require additional processing on raw spectra, and performed well.

Rapid and Non-destructive Classification of New and Aged Maize Seeds Using Hyperspectral Image and Chemometric Methods

The aged seeds have a significant influence on seed vigor and corn growth. Therefore, it is vital for the planting industry to identify aged seeds. In this study, hyperspectral reflectance imaging (1,000-2,000 nm) was employed for identifying aged maize seeds using seeds harvested in different years. The average spectra of the embryo side, endosperm side, and both sides were extracted. The support vector machine (SVM) algorithm was used to develop classification models based on full spectra to evaluate the potential of hyperspectral imaging for maize seed detection and using the principal component analysis (PCA) and ANOVA to reduce data dimensionality and extract feature wavelengths. The classification models achieved perfect performance using full spectra with an accuracy of 100% for the prediction set. The performance of models established with the first three principal components was similar to full spectrum models, but that of PCA loading models was worse. Compared to other spectra, the two-band ratio (1,987 nm/1,079 nm) selected by ANOVA from embryo-side spectra achieved a better classification accuracy of 95% for the prediction set. The image texture features, including histogram statistics (HS) and gray-level co-occurrence matrix (GLCM), were extracted from the two-band ratio image to establish fusion models. The results demonstrated that the two-band ratio selected from embryo-side spectra combined with image texture features achieved the classification of maize seeds harvested in different years with an accuracy of 97.5% for the prediction set. The overall results indicated that combining the two wavelengths with image texture features could detect aged maize seeds effectively. The proposed method was conducive to the development of multi-spectral detection equipment.

Application of chemometrics using direct spectroscopic methods as a QC tool in pharmaceutical industry and their validation

This present review described the application of chemometrics using direct spectroscopic methods at the quality control (QC) laboratory of Pharmaceutical Industries. Using chemometrics methods, all QC assessments during the fabrication processes of the drug preparations can be well performed. Chemometrics methods have some advantages compared to the conventional methods, i.e., non-destructive, can be performed directly to intake samples without any extractions, unnecessary performing stability studies, and cost-effective. To achieve reliable results of analyses, all methods must be validated first prior to routine applications. According to the current Pharmacopeia, the validation parameters are specificity/selectivity, accuracy, repeatability, intermediate precision, range, detection limit, quantification limit and robustness. These validation data must meet the acceptance criteria, that have been described by the analytical target profile (ATP) of the drug preparations.

Variety classification of coated maize seeds based on Raman hyperspectral imaging

As an essential factor in quality assessment of maize seeds, variety purity profoundly impacts final yield and farmers' economic benefits. In this study, a novel method based on Raman hyperspectral imaging system was applied to achieve variety classification of coated maize seeds. A total of 760 maize seeds including 4 different varieties were evaluated. Raman spectral data of 400-1800 cm^-1 were extracted and preprocessed. Variable selection methods involved were modified competitive adaptive reweighted sampling (MCARS), successive projections algorithm (SPA), and their combination. In addition, MCARS was proposed for the first time in this paper as a stable search technology. The performance of support vector machine (SVM) models optimized by genetic algorithm (GA) was analyzed and compared with models based on random forest (RF) and back-propagation neural network (BPNN). Same models based on Vis-NIR spectral data were also established for comparison. Results showed that the MCARS-GA-SVM model based on Raman spectral data obtained the best performance with calibration accuracy of 99.29% and prediction accuracy of 100%, which were stable and easily replicated. In addition, the accuracy on the independent validation set was 96.88%, which proved that the model can be applied in practice. A more simplified MCARS-SPA-GA-SVM model, which contained only 3 variables, had more than 95% accuracy on each data set. This procedure can help to develop a real-time detection system to classify coated seed varieties with high accuracy, which is of great significance for assessing variety purity and increasing crop yield.

Near-infrared hyperspectral imaging technology combined with deep convolutional generative adversarial network to predict oil content of single maize kernel

Rapidly and non-destructively predicting the oil content of single maize kernel is crucial for food industry. However, obtaining a large number of oil content reference values of maize kernels is time-consuming and expensive, and the limited data set also leads to low generalization ability of the model. Here, hyperspectral imaging technology and deep convolutional generative adversarial network (DCGAN) were combined to predict the oil content of single maize kernel. DCGAN was used to simultaneously expand their spectral data and oil content data. After many iterations, fake data that was very similar to the experimental data was generated. Partial least squares regression (PLSR) and support vector regression (SVR) models were established respectively, and their performance was compared before and after data augmentation. The results showed that this method not only improved the performance of two regression models, but also solved the problem of requiring a large amount of training data.

Hyperspectral Imaging Combined With Deep Transfer Learning for Rice Disease Detection

Various rice diseases threaten the growth of rice. It is of great importance to achieve the rapid and accurate detection of rice diseases for precise disease prevention and control. Hyperspectral imaging (HSI) was performed to detect rice leaf diseases in four different varieties of rice. Considering that it costs much time and energy to develop a classifier for each variety of rice, deep transfer learning was firstly introduced to rice disease detection across different rice varieties. Three deep transfer learning methods were adapted for 12 transfer tasks, namely, fine-tuning, deep CORrelation ALignment (CORAL), and deep domain confusion (DDC). A self-designed convolutional neural network (CNN) was set as the basic network of the deep transfer learning methods. Fine-tuning achieved the best transferable performance with an accuracy of over 88% for the test set of the target domain in the majority of transfer tasks. Deep CORAL obtained an accuracy of over 80% in four of all the transfer tasks, which was superior to that of DDC. A multi-task transfer strategy has been explored with good results, indicating the potential of both pair-wise, and multi-task transfers. A saliency map was used for the visualization of the key wavelength range captured by CNN with and without transfer learning. The results indicated that the wavelength range with and without transfer learning was overlapped to some extent. Overall, the results suggested that deep transfer learning methods could perform rice disease detection across different rice varieties. Hyperspectral imaging, in combination with the deep transfer learning method, is a promising possibility for the efficient and cost-saving field detection of rice diseases among different rice varieties.

Application of Visible/Infrared Spectroscopy and Hyperspectral Imaging With Machine Learning Techniques for Identifying Food Varieties and Geographical Origins

Food quality and safety are strongly related to human health. Food quality varies with variety and geographical origin, and food fraud is becoming a threat to domestic and global markets. Visible/infrared spectroscopy and hyperspectral imaging techniques, as rapid and non-destructive analytical methods, have been widely utilized to trace food varieties and geographical origins. In this review, we outline recent research progress on identifying food varieties and geographical origins using visible/infrared spectroscopy and hyperspectral imaging with the help of machine learning techniques. The applications of visible, near-infrared, and mid-infrared spectroscopy as well as hyperspectral imaging techniques on crop food, beverage, fruits, nuts, meat, oil, and some other kinds of food are reviewed. Furthermore, existing challenges and prospects are discussed. In general, the existing machine learning techniques contribute to satisfactory classification results. Follow-up researches of food varieties and geographical origins traceability and development of real-time detection equipment are still in demand.

Combining prior knowledge with input selection algorithms for quantitative analysis using neural networks in laser induced breakdown spectroscopy

Laser-induced breakdown spectroscopy (LIBS) is an emerging technique for the analysis of rocks and mineral samples. Artificial neural networks (ANNs) have been used to estimate the concentration of minerals in samples from LIBS spectra. These spectra are very high dimensional data, and it is known that only specific wavelengths have information on the atomic and molecular features of the sample under investigation. This work presents a systematic methodology based on the Akaike information criterion (AIC) for selecting the wavelengths of LIBS spectra as well as the ANN model complexity, by combining prior knowledge and variable selection algorithms. Several variable selection algorithms are compared within the proposed methodology, namely KBest, a least absolute shrinkage and selection operator (LASSO) regularization, principal component analysis (PCA), and competitive adaptive reweighted sampling (CARS). As an illustrative example, the estimation of copper, iron and arsenic concentrations in pelletized mineral samples is performed. A dataset of LIBS emission spectra with 12 287 wavelengths in the range of 185-1049 nm obtained from 131 samples of copper concentrates is used for regression analysis. An ANN is then trained considering the selected reduced wavelength data. The results are satisfactory using LASSO and CARS algorithms along with prior knowledge, showing that the proposed methodology is very effective for selecting wavelengths and model complexity in quantitative analyses based on ANNs and LIBS.

Hyperspectral imaging coupled with multivariate methods for seed vitality estimation and forecast for Quercus variabilis

In this study, the feasibility of estimation and forecast of different vitality Quercus variabilis seeds by a hyperspectral imaging technique were investigated. Artificially accelerated aging was conducive to achieve the division of four vitality levels. Hyperspectral data in the first 10 h of germination were continuously collected at one-hour intervals. The optimal band was selected for the original and pre-processed spectra which were treated by multiple scatter correction (MSC) and the Savitzky-Golay first derivative (SG 1st). Five characteristic wavelength methods were compared: successive projections algorithm (SPA), competitive adaptive reweighted sampling (CARS), genetic algorithm (GA), variable important in projection (VIP), and random frog (RF). Partial least square-discriminant analysis (PLS-DA) and K-nearest neighbor (KNN) built the vitality estimation model based on different data sets, and GA + PLS-DA constructed the optimal model with the highest accuracy. According to the weight coefficient and reflectance of the characteristic band extracted by the GA, the reflectance curves of different levels over time were plotted. The data of 0 h was employed to establish the vitality forecast model. The forecast model had a high recognition rate, with PLS-DA exceeding 99% and KNN exceeding 85%. This indicated that hyperspectral imaging of seed germination processes could achieve non-destructive estimation of Q. variabilis seed vitality, and accurate prediction in a shorter time is feasible.

Online Application of a Hyperspectral Imaging System for the Sorting of Adulterated Almonds

Almonds are nutrient-rich nuts. Due to their high level of consumption and relatively high price, their production is targeted for illegal practices, with the intention of earning more profit. The most common adulterants are based on superficial matching, and as an adulterant, the apricot kernel is comparatively inexpensive and almost identical in color, texture, odor, and other physicochemical characteristics to almonds. In this study, a near-infrared hyperspectral imaging (NIR-HSI) system in the wavelength range of 900–1700 nm synchronized with a conveyor belt was used for the online detection of added apricot kernels in almonds. A total of 448 samples from different varieties of almonds and apricot kernels (112 × 4) were scanned while the samples moved on the conveyor belt. The spectral data were extracted from each imaged nut and used to develop a partial least square discrimination analysis (PLS-DA) model coupled with different preprocessing techniques. The PLS-DA model displayed over a 97% accuracy for the validation set. Additionally, the beta coefficient obtained from the developed model was used for pixel-based classification. An image processing algorithm was developed for the chemical mapping of almonds and apricot kernels. Consequently, the obtained model was transferred for the online sorting of seeds. The online classification system feedback had an overall accuracy of 85% for the classification of nuts. However, the model presented a relatively low accuracy when evaluated in real-time for online application, which might be due to the rough distribution of samples on the conveyor belt, high speed, delaying time in suction, and lighting variations. Nevertheless, the developed online prototype (NIR-HSI) system combined with multivariate analysis exhibits strong potential for the classification of adulterated almonds, and the results indicate that the system can be effectively used for the high-throughput screening of adulterated almond nuts in an industrial environment.

Identification of Bacterial Blight Resistant Rice Seeds Using Terahertz Imaging and Hyperspectral Imaging Combined With Convolutional Neural Network

Because bacterial blight (BB) disease seriously affects the yield and quality of rice, breeding BB resistant rice is an important priority for plant breeders but the process is time-consuming. The feasibility of using terahertz imaging technology and near-infrared hyperspectral imaging technology to identify BB resistant seeds has therefore been studied. The two-dimensional (2D) spectral images and one-dimensional (1D) spectra provided by both imaging methods were used to build discriminant models based on a deep learning method, the convolutional neural network (CNN), and traditional machine learning methods, support vector machine (SVM), random forest (RF), and partial least squares discriminant analysis (PLS-DA). The highest classification accuracy was achieved by the discriminate model based on CNN using the terahertz absorption spectra. Confusion matrixes were pictured to show the identification details. The t-distributed stochastic neighbor embedding (t-SNE) method was used to visualize the process of CNN data processing. Terahertz imaging technology combined with CNN has great potential to quickly identify BB resistant rice seeds and is more accurate than using near-infrared hyperspectral imaging.

Hyperspectral imaging for accurate determination of rice variety using a deep learning network with multi-feature fusion

The phenomena of rice adulteration and shoddy rice arise continuously in high-quality rice and reduce the interests of producers, consumers and traders. Hyperspectral imaging (HSI) was conducted to determine rice variety using a deep learning network with multiple features, namely, spectroscopy, texture and morphology. HSI images of 10 representative high-quality rice varieties in China were measured. Spectroscopy and morphology were extracted from HSI images and binary images in region of interest, respectively. And texture was obtained from the monochromatic images of characteristic wavelengths which were highly correlated with rice varieties. A deep learning network, namely principal component analysis network (PCANet), was adopted with these features to develop classification models for determining rice variety, and machine learning methods as K-nearest neighbour and random forest were used to compare with PCANet. Meanwhile, multivariate scatter correction, standard normal variate, Savitzky-Golay smoothing and Savitzky-Golay's first-order were applied to eliminate spectral interference, and principal component analysis (PCA) was performed to obtain the main information of high-dimensional features. Multi-feature fusion improved recognition accuracy, and PCANet demonstrated considerable advantage in classification performance. The best result was achieved by PCANet with PCA-processed spectroscopic and texture features with correct classification rates of 98.66% and 98.57% for the training and prediction sets, respectively. In summary, the proposed method provides an accurate identification of rice variety and can be easily extended to the classification, attribution and grading of other agricultural products.

A modified feature fusion method for distinguishing seed strains using hyperspectral data

Precise classification of seeds is important for agriculture. Due to the slight physical and chemical difference between different types of wheat and high correlation between bands of images, it is easy to fall into the local optimum when selecting the characteristic band of using the spectral average only. In this paper, in order to solve this problem, a new variable fusion strategy was proposed based on successive projection algorithm and the variable importance in projection algorithm to obtain a comprehensive and representative variable feature for higher classification accuracy, within spectral mean and spectral standard deviation, so the 25 feature bands obtained are classified by support vector machine, and the classification accuracy rate reached 83.3%. It indicates that the new fusion strategy can mine the effective features of hyperspectral data better to improve the accuracy of the model and it can provide a theoretical basis for the hyperspectral classification of tiny kernels.

Application of near-infrared hyperspectral imaging to identify a variety of silage maize seeds and common maize seeds

Common maize seeds and silage maize seeds are similar in appearance and are difficult to identify with the naked eye. Four varieties of common maize seeds and four varieties of silage maize seeds were identified by near-infrared hyperspectral imaging (NIR-HSI) combined with chemometrics. The pixel-wise principal component analysis was used to distinguish the differences among different varieties of maize seeds. The object-wise spectra of each single seed sample were extracted to build classification models. Support vector machine (SVM) and radial basis function neural network (RBFNN) classification models were established using two different classification strategies. First, the maize seeds were directly classified into eight varieties with the prediction accuracy of the SVM model and RBFNN model over 86%. Second, the seeds of silage maize and common maize were firstly classified with the classification accuracy over 88%, then the seeds were classified into four varieties, respectively. The classification accuracy of silage maize seeds was over 98%, and the classification accuracy of common maize seeds was over 97%. The results showed that the varieties of common maize seeds and silage maize seeds could be classified by NIR-HSI combined with chemometrics, which provided an effective means to ensure the purity of maize seeds, especially to isolate common seeds and silage seeds.

Hyperspectral imaging technology combined with deep forest model to identify frost-damaged rice seeds

In recent years, deep learning models have been widely used in the field of hyperspectral imaging. However, the training of deep learning models requires not only a large number of samples, but also the need to set too many hyper-parameters, which is time consuming and laborious for researchers. This study used hyperspectral imaging technology combined with a deep learning model suitable for small-scale sample data sets, deep forests (DF) model, to classify rice seeds with different degrees of frost damage. During the period, three spectral preprocessing methods (Savitzky-Golay first derivative (SG1), standard normal variate (SNV), and multivariate scatter correction (MSC)) were used to process the original spectral data, and three feature extraction algorithms (principal component analysis (PCA), successive projections algorithm (SPA), and neighborhood component analysis (NCA)) were used to extract the characteristic wavelengths. Moreover, DF model and three traditional machine learning models (decision tree (DT), k-nearest neighbor (KNN), and support vector machine (SVM)) were built based on different numbers of sample sets. After multivariate data analysis, it showed that the pretreatment effect of MSC was the most excellent, and the characteristic wavelength extracted by NCA algorithm was the most useful. In addition, the performance of DF model was better than these three traditional classifier models, and it still performed well in small-scale sample set data. Therefore, DF model was chosen as the best classification model. The results of this study show that the DF model maintains good classification performance in small-scale sample set data, and it has a good application prospect in hyperspectral imaging technology.

A Rapid and Highly Efficient Method for the Identification of Soybean Seed Varieties: Hyperspectral Images Combined with Transfer Learning

Convolutional neural network (CNN) can be used to quickly identify crop seed varieties. 1200 seeds of ten soybean varieties were selected, hyperspectral images of both the front and the back of the seeds were collected, and the reflectance of soybean was derived from the hyperspectral images. A total of 9600 images were obtained after data augmentation, and the images were divided into a training set, validation set, and test set with a 3:1:1 ratio. Pretrained models (AlexNet, ResNet18, Xception, InceptionV3, DenseNet201, and NASNetLarge) after fine-tuning were used for transfer training. The optimal CNN model for soybean seed variety identification was selected. Furthermore, the traditional machine learning models for soybean seed variety identification were established by using reflectance as input. The results show that the six models all achieved 91% accuracy in the validation set and achieved accuracy values of 90.6%, 94.5%, 95.4%, 95.6%, 96.8%, and 97.2%, respectively, in the test set. This method is better than the identification of soybean seed varieties based on hyperspectral reflectance. The experimental results support a novel method for identifying soybean seeds rapidly and accurately, and this method also provides a good reference for the identification of other crop seeds.

Rapid Classification of Wheat Grain Varieties Using Hyperspectral Imaging and Chemometrics

The classification of wheat grain varieties is of great value because its high purity is the yield and quality guarantee. In this study, hyperspectral imaging combined with the chemometric methods was applied to explore and implement the varieties classification of wheat seeds. The hyperspectral images of all the samples covering 874–1734 nm bands were collected. Exploratory analysis was first carried out while using principal component analysis (PCA) and linear discrimination analysis (LDA). Spectral preprocessing methods including standard normal variate (SNV), multiplicative scatter correction (MSC), and wavelet transform (WT) were introduced, and their effects on discriminant models were studied to eliminate the interference of instrumental and environmental factors. PCA loading, successive projections algorithm (SPA), and random frog (RF) were applied to extract feature wavelengths for redundancy elimination owing to the possibility of existing redundant spectral information. Classification models were developed based on full wavelengths and feature wavelengths using LDA, support vector machine (SVM), and extreme learning machine (ELM). This optimal model was finally utilized to generate visualization map to observe the classification performance intuitively. When comparing with other models, ELM based on full wavelengths achieved the best accuracy up to 91.3%. The overall results suggested that hyperspectral imaging was a potential tool for the rapid and accurate identification of wheat varieties, which could be conducted in large-scale seeds classification and quality detection in modern seed industry.

Identification of Soybean Varieties Using Hyperspectral Imaging Coupled with Convolutional Neural Network

Soybean variety is connected to stress resistance ability, as well as nutritional and commercial value. Near-infrared hyperspectral imaging was applied to classify three varieties of soybeans (Zhonghuang37, Zhonghuang41, and Zhonghuang55). Pixel-wise spectra were extracted and preprocessed, and average spectra were also obtained. Convolutional neural networks (CNN) using the average spectra and pixel-wise spectra of different numbers of soybeans were built. Pixel-wise CNN models obtained good performance predicting pixel-wise spectra and average spectra. With the increase of soybean numbers, performances were improved, with the classification accuracy of each variety over 90%. Traditionally, the number of samples used for modeling is large. It is time-consuming and requires labor to obtain hyperspectral data from large batches of samples. To explore the possibility of achieving decent identification results with few samples, a majority vote was also applied to the pixel-wise CNN models to identify a single soybean variety. Prediction maps were obtained to present the classification results intuitively. Models using pixel-wise spectra of 60 soybeans showed equivalent performance to those using the average spectra of 810 soybeans, illustrating the possibility of discriminating soybean varieties using few samples by acquiring pixel-wise spectra.

Hyperspectral imaging for seed quality and safety inspection: a review

Hyperspectral imaging has attracted great attention as a non-destructive and fast method for seed quality and safety assessment in recent years. The capability of this technique for classification and grading, viability and vigor detection, damage (defect and fungus) detection, cleanness detection and seed composition determination is illustrated by presentation of applications in quality and safety determination of seed in this review. The summary of hyperspectral imaging technology for seed quality and safety inspection for each category is also presented, including the analyzed spectral range, sample varieties, sample status, sample numbers, features (spectral features, image features, feature extraction methods), signal mode and data analysis strategies. The successful application of hyperspectral imaging in seed quality and safety inspection proves that many routine seed inspection tasks can be facilitated with hyperspectral imaging.

Cultivar Classification of Single Sweet Corn Seed Using Fourier Transform Near-Infrared Spectroscopy Combined with Discriminant Analysis

Seed purity is a key indicator of crop seed quality. The conventional methods for cultivar identification are time-consuming, expensive, and destructive. Fourier transform near-infrared (FT-NIR) spectroscopy combined with discriminant analyses, was studied as a rapid and nondestructive technique to classify the cultivars of sweet corn seeds. Spectra with a range of 1000–2500 nm collected from 760 seeds of two cultivars were used for the discriminant analyses. Thereafter, 126 feature wavelengths were identified from 1557 wavelengths using a genetic algorithm (GA) to build simplified classification models. Four classification algorithms, namely K-nearest neighbor (KNN), soft independent method of class analogy (SIMCA), partial least-squares discriminant analysis (PLS-DA), and support vector machine discriminant analysis (SVM-DA) were tested on full-range wavelengths and feature wavelengths, respectively. With the full-range wavelengths, all four algorithms achieved a high classification accuracy range from 97.56% to 99.59%, and the SVM-DA worked better than other models. From the feature wavelengths, no significant decline in accuracies was observed in most of the models and a high accuracy of 99.19% was still obtained by the PLS-DA model. This study demonstrated that using the FT-NIR technique with discriminant analyses could be a feasible way to classify sweet corn seed cultivars and the proper classification model could be embedded in seed sorting machinery to select high-purity seeds.

Rapid prediction of chlorophylls and carotenoids content in tea leaves under different levels of nitrogen application based on hyperspectral imaging

BACKGROUND

Photosynthetic pigments perform critical physiological functions in tea plants. Their content is an essential indicator of photosynthetic efficiency and nutritional status. The present study aimed to predict chlorophyll a (Chl a), chlorophyll b (Chl b), total chlorophyll (total Chl), and carotenoid (Car) content in tea leaves under different levels of nitrogen treatment using hyperspectral imaging (HSI) in combination with variable selection algorithms.

RESULTS

A total of 150 samples were collected and scanned using the HSI system. The mean spectrum in the region of interest (ROI) was extracted, and the pigment content was measured by traditional chemical methods. Five and seven optimal wavelengths (OWs) were selected using the regression coefficients (RCs) of partial least squares regression (PLSR) and the second-derivative (2-Der), respectively. The optimal 2-Der-PLSR models for Chl a, Chl b, total Chl, and Car performed remarkably well based on seven OWs with correlation coefficients of prediction (R_P ) of 0.9337, 0.9322, 0.9333 and 0.9036, root mean square errors in prediction (RMSEP) of 0.1100, 0.0511, 0.1620, and 0.0300 mg g^-1 , respectively.

CONCLUSION

The results of this study revealed that HSI combined with variable selection method can be employed as a rapid and accurate method for predicting the content of pigments in tea plants. © 2018 Society of Chemical Industry.

Classification of Frozen Corn Seeds Using Hyperspectral VIS/NIR Reflectence Imaging

A VIS/NIR hyperspectral imaging system was used to classify three different degrees of freeze-damage in corn seeds. Using image processing methods, the hyperspectral image of the corn seed embryo was obtained first. To find a relatively better method for later imaging visualization, four different pretreatment methods (no pretreatment, multiplicative scatter correction (MSC), standard normal variation (SNV) and 5 points and 3 times smoothing (5-3 smoothing)), four wavelength selection algorithms (successive projection algorithm (SPA), principal component analysis (PCA), X-loading and full-band method) and three different classification modeling methods (partial least squares-discriminant analysis (PLS-DA), K-nearest neighbor (KNN) and support vector machine (SVM)) were applied to make a comparison. Next, the visualization images according to a mean spectrum to mean spectrum (M2M) and a mean spectrum to pixel spectrum (M2P) were compared in order to better represent the freeze damage to the seed embryos. It was concluded that the 5-3 smoothing method and SPA wavelength selection method applied to the modeling can improve the signal-to-noise ratio, classification accuracy of the model (more than 90%). The final classification results of the method M2P were better than the method M2M, which had fewer numbers of misclassified corn seed samples and the samples could be visualized well.

Variety identification of oat seeds using hyperspectral imaging: investigating the representation ability of deep convolutional neural network

Variety identification of seeds is critical for assessing variety purity and ensuring crop yield. In this paper, a novel method based on hyperspectral imaging (HSI) and deep convolutional neural network (DCNN) was proposed to discriminate the varieties of oat seeds. The representation ability of DCNN was also investigated. The hyperspectral images with a spectral range of 874–1734 nm were primarily processed by principal component analysis (PCA) for exploratory visual distinguishing. Then a DCNN trained in an end-to-end manner was developed. The deep spectral features automatically learnt by DCNN were extracted and combined with traditional classifiers (logistic regression (LR), support vector machine with RBF kernel (RBF_SVM) and linear kernel (LINEAR_SVM)) to construct discriminant models. Contrast models were built based on the traditional classifiers using full wavelengths and optimal wavelengths selected by the second derivative (2nd derivative) method. The comparison results showed that all DCNN-based models outperformed the contrast models. DCNN trained in an end-to-end manner achieved the highest accuracy of 99.19% on the testing set, which was finally employed to visualize the variety classification. The results demonstrated that the deep spectral features with outstanding representation ability enabled HSI together with DCNN to be a reliable tool for rapid and accurate variety identification, which would help to develop an on-line system for quality detection of oat seeds as well as other grain seeds.

The excellent representation ability of deep spectral features enables hyperspectral imaging combined with deep convolutional neural network to be a powerful tool for large-scale seeds detection in modern seed industry.

Identification of Hybrid Okra Seeds Based on Near-Infrared Hyperspectral Imaging Technology

Near-infrared (874–1734 nm) hyperspectral imaging technology combined with chemometrics was used to identify parental and hybrid okra seeds. A total of 1740 okra seeds of three different varieties, which contained the male parent xiaolusi, the female parent xianzhi, and the hybrid seed penzai, were collected, and all of the samples were randomly divided into the calibration set and the prediction set in a ratio of 2:1. Principal component analysis (PCA) was applied to explore the separability of different seeds based on the spectral characteristics of okra seeds. Fourteen and 86 characteristic wavelengths were extracted by using the successive projection algorithm (SPA) and competitive adaptive reweighted sampling (CARS), respectively. Another 14 characteristic wavelengths were extracted by using CARS combined with SPA. Partial least squares discriminant analysis (PLS-DA) and support vector machine (SVM) were developed based on the characteristic wavelength and full-band spectroscopy. The experimental results showed that the SVM discriminant model worked well and that the correct recognition rate was over 93.62% based on full-band spectroscopy. As for the discriminative model that was based on characteristic wavelength, the SVM model based on the CARS algorithm was better than the other two models. Combining the CARS+SVM calibration model and image processing technology, a pseudo-color map of sample prediction was generated, which could intuitively identify the species of okra seeds. The whole process provided a new idea for agricultural breeding in the rapid screening and identification of hybrid okra seeds.

Non-Destructive and Rapid Variety Discrimination and Visualization of Single Grape Seed Using Near-Infrared Hyperspectral Imaging Technique and Multivariate Analysis

Hyperspectral images in the spectral range of 874⁻1734 nm were collected for 14,015, 14,300 and 15,042 grape seeds of three varieties, respectively. Pixel-wise spectra were preprocessed by wavelet transform, and then, spectra of each single grape seed were extracted. Principal component analysis (PCA) was conducted on the hyperspectral images. Scores for images of the first six principal components (PCs) were used to qualitatively recognize the patterns among different varieties. Loadings of the first six PCs were used to identify the effective wavelengths (EWs). Support vector machine (SVM) was used to build the discriminant model using the spectra based on the EWs. The results indicated that the variety of each single grape seed was accurately identified with a calibration accuracy of 94.3% and a prediction accuracy of 88.7%. An external validation image of each variety was used to evaluate the proposed model and to form the classification maps where each single grape seed was explicitly identified as belonging to a distinct variety. The overall results indicated that a hyperspectral imaging (HSI) technique combined with multivariate analysis could be used as an effective tool for non-destructive and rapid variety discrimination and visualization of grape seeds. The proposed method showed great potential for developing a multi-spectral imaging system for practical application in the future.