Near-Infrared Hyperspectral Imaging Combined with Deep Learning to Identify Cotton Seed Varieties

Cotton seed cultivar identification based on the fusion of spectral and textural features

The mixing of cotton seeds of different cultivars and qualities can lead to differences in growth conditions and make field management difficult. In particular, except for yield loss, it can also lead to inconsistent cotton quality and poor textile product quality, causing huge economic losses to farmers and the cotton processing industry. However, traditional cultivar identification methods for cotton seeds are time-consuming, labor-intensive, and cumbersome, which cannot meet the needs of modern agriculture and modern cotton processing industry. Therefore, there is an urgent need for a fast, accurate, and non-destructive method for identifying cotton seed cultivars. In this study, hyperspectral images (397.32 nm-1003.58 nm) of five cotton cultivars, namely Jinke 20, Jinke 21, Xinluzao 64, Xinluzao 74, and Zhongmiansuo 5, were captured using a Specim IQ camera, and then the average spectral information of seeds of each cultivar was used for spectral analysis, aiming to estab-lish a cotton seed cultivar identification model. Due to the presence of many obvious noises in the < 400 nm and > 1000 nm regions of the collected spectral data, spectra from 400 nm to 1000 nm were selected as the representative spectra of the seed samples. Then, various denoising techniques, including Savitzky-Golay (SG), Standard Normal Variate (SNV), and First Derivative (FD), were applied individually and in combination to improve the quality of the spectra. Additionally, a successive projections algorithm (SPA) was employed for spectral feature selection. Based on the full-band spectra, a Partial Least Squares-Discriminant Analysis (PLS-DA) model was established. Furthermore, spectral features and textural features were fused to create Random Forest (RF), Convolutional Neural Network (CNN), and Extreme Learning Machine (ELM) identification models. The results showed that: (1) The SNV-FD preprocessing method showed the optimal denoising performance. (2) SPA highlighted the near-infrared region (800-1000 nm), red region (620-700 nm), and blue-green region (420-570 nm) for identifying cotton cultivar. (3) The fusion of spectral features and textural features did not consistently improve the accuracy of all modeling strategies, suggesting the need for further research on appropriate modeling strategies. (4) The ELM model had the highest cotton cultivar identification accuracy, with an accuracy of 100% for the training set and 98.89% for the test set. In conclusion, this study successfully developed a highly accurate cotton seed cultivar identification model (ELM model). This study provides a new method for the rapid and non-destructive identification of cotton seed cultivars, which will help ensure the cultivar consistency of seeds used in cotton planting, and improve the overall quality and yield of cotton.

A maize seed variety identification method based on improving deep residual convolutional network

Seed quality and safety are related to national food security, and seed variety purity is an essential indicator in seed quality detection. This study established a maize seed dataset comprising 5877 images of six different types and proposed a maize seed recognition model based on an improved ResNet50 framework. Firstly, we introduced the ResStage structure in the early stage of the original model, which facilitated the network's learning process and enabled more efficient information propagation across the network layers. Meanwhile, in the later residual blocks of the model, we introduced both the efficient channel attention (ECA) mechanism and depthwise separable (DS) convolution, which reduced the model's parameter cost and enabled the capturing of more precise and detailed features. Finally, a Swish-PReLU mixed activation function was introduced globally to improve the overall predictive power of the model. The results showed that our model achieved an impressive accuracy of 91.23% in corn seed classification, surpassing other related models. Compared with the original model, our model improved the accuracy by 7.07%, reduced the loss value by 0.19, and decreased the number of parameters by 40%. The research suggested that this method can efficiently classify corn seeds, holding significant value in seed variety identification.

HairNet2: deep learning to quantify cotton leaf hairiness, a complex genetic and environmental trait

BACKGROUND

Cotton accounts for 80% of the global natural fibre production. Its leaf hairiness affects insect resistance, fibre yield, and economic value. However, this phenotype is still qualitatively assessed by visually attributing a Genotype Hairiness Score (GHS) to a leaf/plant, or by using the HairNet deep-learning model which also outputs a GHS. Here, we introduce HairNet2, a quantitative deep-learning model which detects leaf hairs (trichomes) from images and outputs a segmentation mask and a Leaf Trichome Score (LTS).

RESULTS

Trichomes of 1250 images were annotated (AnnCoT) and a combination of six Feature Extractor modules and five Segmentation modules were tested alongside a range of loss functions and data augmentation techniques. HairNet2 was further validated on the dataset used to build HairNet (CotLeaf-1), a similar dataset collected in two subsequent seasons (CotLeaf-2), and a dataset collected on two genetically diverse populations (CotLeaf-X). The main findings of this study are that (1) leaf number, environment and image position did not significantly affect results, (2) although GHS and LTS mostly correlated for individual GHS classes, results at the genotype level revealed a strong LTS heterogeneity within a given GHS class, (3) LTS correlated strongly with expert scoring of individual images.

CONCLUSIONS

HairNet2 is the first quantitative and scalable deep-learning model able to measure leaf hairiness. Results obtained with HairNet2 concur with the qualitative values used by breeders at both extremes of the scale (GHS 1-2, and 5-5+), but interestingly suggest a reordering of genotypes with intermediate values (GHS 3-4+). Finely ranking mild phenotypes is a difficult task for humans. In addition to providing assistance with this task, HairNet2 opens the door to selecting plants with specific leaf hairiness characteristics which may be associated with other beneficial traits to deliver better varieties.

Remote Sensing Techniques: Hyperspectral Imaging and Data Analysis

Hyperspectral imaging is a remote sensing technique that enables remote, noninvasive measurement of plant traits. Here, we outline the procedures for camera setup, scanning, and calibration, along with the acquisition of black and white reference materials, which are the key steps in collecting hyperspectral imagery. We also discuss the development of predictive models such as partial least-squares regression, using both large and small datasets, which are used to predict plant traits from hyperspectral data. To ensure practical applicability, we provide code examples that allow readers to immediately implement these techniques in real-world scenarios. We introduce these topics to beginners in an accessible and understandable manner.

Integrating spectral and image information for prediction of cottonseed vitality

Cotton plays a significant role in people's lives, and cottonseeds serve as a vital assurance for successful cotton cultivation and production. Premium-quality cottonseeds can significantly enhance the germination rate of cottonseeds, resulting in increased cotton yields. The vitality of cottonseeds is a crucial metric that reflects the quality of the seeds. However, currently, the industry lacks a non-destructive method to directly assess cottonseed vitality without compromising the integrity of the seeds. To address this challenge, this study employed a hyperspectral imaging acquisition system to gather hyperspectral data on cottonseeds. This system enables the simultaneous collection of hyperspectral data from 25 cottonseeds. This study extracted spectral and image information from the hyperspectral data of cottonseeds to predict their vitality. SG, SNV, and MSC methods were utilized to preprocess the spectral data of cottonseeds. Following this preprocessing step, feature wavelength points of the cottonseeds were extracted using SPA and CARS algorithms. Subsequently, GLCM was employed to extract texture features from images corresponding to these feature wavelength points, including attributes such as Contrast, Correlation, Energy, and Entropy. Finally, the vitality of cottonseeds was predicted using PLSR, SVR, and a self-built 1D-CNN model. For spectral data analysis, the 1D-CNN model constructed after MSC+CARS preprocessing demonstrated the highest performance, achieving a test set correlation coefficient of 0.9214 and an RMSE of 0.7017. For image data analysis, the 1D-CNN model constructed after SG+CARS preprocessing outperformed the others, yielding a test set correlation coefficient of 0.8032 and an RMSE of 0.9683. In the case of fused spectral and image data, the 1D-CNN model built after SG+SPA preprocessing displayed the best performance, attaining a test set correlation coefficient of 0.9427 and an RMSE of 0.6872. These findings highlight the effectiveness of the 1D-CNN model and the fusion of spectral and image features for cottonseed vitality prediction. This research contributes significantly to the development of automated detection devices for assessing cottonseed vitality.

Intelligent identification on cotton verticillium wilt based on spectral and image feature fusion

BACKGROUND

Verticillium wilt is the major disease of cotton, which would cause serious yield reduction and economic losses, and the identification of cotton verticillium wilt is of great significance to cotton research. However, the traditional method is still manual, which is subjective, inefficient, and labor-intensive, and therefore, this study has proposed a novel method for cotton verticillium wilt identification based on spectral and image feature fusion. The cotton hyper-spectral images have been collected, while the regions of interest (ROI) have been extracted as samples including 499 healthy leaves and 498 diseased leaves, and the average spectral information and RGB image of each sample were obtained. In spectral feature processing, the preprocessing methods including Savitzky-Golay smoothing (SG), multiplicative scatter correction (MSC), de-trending (DT) and mean normalization (MN) algorithms have been adopted, while the feature band extraction methods have adopted principal component analysis (PCA) and successive projections algorithm (SPA). In RGB image feature processing, the EfficientNet was applied to build classification model and 16 image features have been extracted from the last convolutional layer. And then, the obtained spectral and image features were fused, while the classification model was established by support vector machine (SVM) and back propagation neural network (BPNN). Additionally, the spectral full bands and feature bands were used as comparison for SVM and BPNN classification respectively.

RESULT

The results showed that the average accuracy of EfficientNet for cotton verticillium wilt identification was 93.00%. By spectral full bands, SG-MSC-BPNN model obtained the better performance with classification accuracy of 93.78%. By feature bands, SG-MN-SPA-BPNN model obtained the better performance with classification accuracy of 93.78%. By spectral and image fused features, SG-MN-SPA-FF-BPNN model obtained the best performance with classification accuracy of 98.99%.

CONCLUSIONS

The study demonstrated that it was feasible and effective to use fused spectral and image features based on hyper-spectral imaging to improve identification accuracy of cotton verticillium wilt. The study provided theoretical basis and methods for non-destructive and accurate identification of cotton verticillium wilt.

A method for detecting the quality of cotton seeds based on an improved ResNet50 model

The accurate and rapid detection of cotton seed quality is crucial for safeguarding cotton cultivation. To increase the accuracy and efficiency of cotton seed detection, a deep learning model, which was called the improved ResNet50 (Impro-ResNet50), was used to detect cotton seed quality. First, the convolutional block attention module (CBAM) was embedded into the ResNet50 model to allow the model to learn both the vital channel information and spatial location information of the image, thereby enhancing the model's feature extraction capability and robustness. The model's fully connected layer was then modified to accommodate the cotton seed quality detection task. An improved LRelu-Softplus activation function was implemented to facilitate the rapid and straightforward quantification of the model training procedure. Transfer learning and the Adam optimization algorithm were used to train the model to reduce the number of parameters and accelerate the model's convergence. Finally, 4419 images of cotton seeds were collected for training models under controlled conditions. Experimental results demonstrated that the Impro-ResNet50 model could achieve an average detection accuracy of 97.23% and process a single image in 0.11s. Compared with Squeeze-and-Excitation Networks (SE) and Coordination Attention (CA), the model's feature extraction capability was superior. At the same time, compared with classical models such as AlexNet, VGG16, GoogLeNet, EfficientNet, and ResNet18, this model had superior detection accuracy and complexity balances. The results indicate that the Impro-ResNet50 model has a high detection accuracy and a short recognition time, which meet the requirements for accurate and rapid detection of cotton seed quality.

Single Seed Near-Infrared Hyperspectral Imaging for Classification of Perennial Ryegrass Seed

The detection of beneficial microbes living within perennial ryegrass seed causing no apparent defects is challenging, even with the most sensitive and conventional methods, such as DNA genotyping. Using a near-infrared hyperspectral imaging system (NIR-HSI), we were able to discriminate not only the presence of the commercial NEA12 fungal endophyte strain but perennial ryegrass cultivars of diverse seed age and batch. A total of 288 wavebands were extracted for individual seeds from hyperspectral images. The optimal pre-processing methods investigated yielded the best partial least squares discriminant analysis (PLS-DA) classification model to discriminate NEA12 and without endophyte (WE) perennial ryegrass seed with a classification accuracy of 89%. Effective wavelength (EW) selection based on GA-PLS-DA resulted in the selection of 75 wavebands yielding 88.3% discrimination accuracy using PLS-DA. For cultivar identification, the artificial neural network discriminant analysis (ANN-DA) was the best-performing classification model, resulting in >90% classification accuracy for Trojan, Alto, Rohan, Governor and Bronsyn. EW selection using GA-PLS-DA resulted in 87 wavebands, and the PLS-DA model performed the best, with no extensive compromise in performance, resulting in >89.1% accuracy. The study demonstrates the use of NIR-HSI reflectance data to discriminate, for the first time, an associated beneficial fungal endophyte and five cultivars of perennial ryegrass seed, irrespective of seed age and batch. Furthermore, the negligible effects on the classification errors using EW selection improve the capability and deployment of optimized methods for real-time analysis, such as the use of low-cost multispectral sensors for single seed analysis and automated seed sorting devices.

Dissection of Hyperspectral Reflectance to Estimate Photosynthetic Characteristics in Upland Cotton (Gossypium hirsutum L.) under Different Nitrogen Fertilizer Application Based on Machine Learning Algorithms

Hyperspectral technology has enabled rapid and efficient nitrogen monitoring in crops. However, most approaches involve direct monitoring of nitrogen content or physiological and biochemical indicators directly related to nitrogen, which cannot reflect the overall plant nutritional status. Two important photosynthetic traits, the fraction of absorbed photosynthetically active radiation (FAPAR) and the net photosynthetic rate (Pn), were previously shown to respond positively to nitrogen changes. Here, Pn and FAPAR were used for correlation analysis with hyperspectral data to establish a relationship between nitrogen status and hyperspectral characteristics through photosynthetic traits. Using principal component and band autocorrelation analyses of the original spectral reflectance, two band positions (350-450 and 600-750 nm) sensitive to nitrogen changes were obtained. The performances of four machine learning algorithm models based on six forms of hyperspectral transformations showed that the light gradient boosting machine (LightGBM) model based on the hyperspectral first derivative could better invert the Pn of function-leaves in cotton, and the random forest (RF) model based on hyperspectral first derivative could better invert the FAPAR of the cotton canopy. These results provide advanced metrics for non-destructive tracking of cotton nitrogen status, which can be used to diagnose nitrogen nutrition and cotton growth status in large farms.

A Block-Based and Highly Parallel CNN Accelerator for Seed Sorting

Seed sorting is critical for the breeding industry to improve the agricultural yield. The seed sorting methods based on convolutional neural networks (CNNs) have achieved excellent recognition accuracy on large-scale pretrained network models. However, CNN inference is a computationally intensive process that often requires hardware acceleration to operate in real time. For embedded devices, the high-power consumption of graphics processing units (GPUs) is generally prohibitive, and the field programmable gate array (FPGA) becomes a solution to perform high-speed inference by providing a customized accelerator for a particular user. To date, the recognition speeds of the FPGA-based universal accelerators for high-throughput seed sorting tasks are slow, which cannot guarantee real-time seed sorting. Therefore, a block-based and highly parallel MobileNetV2 accelerator is proposed in this paper. First, a hardware-friendly quantization method that uses only fixed-point operation is designed to reduce resource consumption. Then, the block convolution strategy is proposed to avoid latency and energy consumption increase caused by large-scale intermediate result off-chip data transfers. Finally, two scalable computing engines are explicitly designed for depth-wise convolution (DWC) and point-wise convolution (PWC) to develop the high parallelism of block convolution computation. Moreover, an efficient memory system with a double buffering mechanism and new data reordering mode is designed to address the imbalance between memory access and parallel computing. Our proposed FPGA-based MobileNetV2 accelerator for real-time seed sorting is implemented and evaluated on the platform of Xilinx XC7020. Experimental results demonstrate that our implementation can achieve about 29.4 frames per second (FPS) and 10.86 Giga operations per second (GOPS), and 0.92× to 5.70 × DSP-efficiency compared with previous FPGA-based accelerators.

Classification of Toona sinensis Young Leaves Using Machine Learning and UAV-Borne Hyperspectral Imagery

Rapid and accurate distinction between young and old leaves of Toona sinensis in the wild is of great significance to the selection of T. sinensis varieties and the evaluation of relative yield. In this study, UAV hyperspectral imaging technology was used to obtain canopy hyperspectral data of biennial seedlings of different varieties of T. sinensis to distinguish young and old leaves. Five classification models were trained, namely Random Forest (RF), Artificial Neural Network (ANN), Decision Tree (DT), Partial Least Squares Discriminant Analysis (PLSDA), and Support Vector Machine (SVM). Raw spectra and six preprocessing methods were used to fit the best classification model. Satisfactory accuracy was obtained from all the five models using the raw spectra. The SVM model showed good performance on raw spectra and all preprocessing methods, and yielded higher accuracy, sensitivity, precision, and specificity than other models. In the end, the SVM model based on the raw spectra produced the most reliable and robust prediction results (99.62% accuracy and 99.23% sensitivity on the validation set only, and 100.00% for the rest). Three important spectral regions of 422.7~503.2, 549.2, and 646.2~687.2 nm were found to be highly correlated with the identification of young leaves of T. sinensis. In this study, a fast and effective method for identifying young leaves of T. sinensis was found, which provided a reference for the rapid identification of young leaves of T. sinensis in the wild.

A model for genuineness detection in genetically and phenotypically similar maize variety seeds based on hyperspectral imaging and machine learning

BACKGROUND

Variety genuineness and purity are essential indices of maize seed quality that affect yield. However, detection methods for variety genuineness are time-consuming, expensive, require extensive training, or destroy the seeds in the process. Here, we present an accurate, high-throughput, cost-effective, and non-destructive method for screening variety genuineness that uses seed phenotype data with machine learning to distinguish between genetically and phenotypically similar seed varieties. Specifically, we obtained image data of seed morphology and hyperspectral reflectance for Jingke 968 and nine other closely-related varieties (non-Jingke 968). We then compared the robustness of three common machine learning algorithms in distinguishing these varieties based on the phenotypic imaging data.

RESULTS

Our results showed that hyperspectral imaging (HSI) combined with a multilayer perceptron (MLP) or support vector machine (SVM) model could distinguish Jingke 968 from varieties that differed by as few as two loci, with a 99% or higher accuracy, while machine vision imaging provided  ~ 90% accuracy. Through model validation and updating with varieties not included in the training data, we developed a genuineness detection model for Jingke 968 that effectively discriminated between genetically similar and distant varieties.

CONCLUSIONS

This strategy has potential for wide adoption in large-scale variety genuineness detection operations for internal quality control or governmental regulatory agencies, or for accelerating the breeding of new varieties. Besides, it could easily be extended to other target varieties and other crops.

Cotton Breeding in Australia: Meeting the Challenges of the 21st Century

The Commonwealth Scientific and Industrial Research Organisation (CSIRO) cotton breeding program is the sole breeding effort for cotton in Australia, developing high performing cultivars for the local industry which is worth∼AU$3 billion per annum. The program is supported by Cotton Breeding Australia, a Joint Venture between CSIRO and the program's commercial partner, Cotton Seed Distributors Ltd. (CSD). While the Australian industry is the focus, CSIRO cultivars have global impact in North America, South America, and Europe. The program is unique compared with many other public and commercial breeding programs because it focuses on diverse and integrated research with commercial outcomes. It represents the full research pipeline, supporting extensive long-term fundamental molecular research; native and genetically modified (GM) trait development; germplasm enhancement focused on yield and fiber quality improvements; integration of third-party GM traits; all culminating in the release of new commercial cultivars. This review presents evidence of past breeding successes and outlines current breeding efforts, in the areas of yield and fiber quality improvement, as well as the development of germplasm that is resistant to pests, diseases and abiotic stressors. The success of the program is based on the development of superior germplasm largely through field phenotyping, together with strong commercial partnerships with CSD and Bayer CropScience. These relationships assist in having a shared focus and ensuring commercial impact is maintained, while also providing access to markets, traits, and technology. The historical successes, current foci and future requirements of the CSIRO cotton breeding program have been used to develop a framework designed to augment our breeding system for the future. This will focus on utilizing emerging technologies from the genome to phenome, as well as a panomics approach with data management and integration to develop, test and incorporate new technologies into a breeding program. In addition to streamlining the breeding pipeline for increased genetic gain, this technology will increase the speed of trait and marker identification for use in genome editing, genomic selection and molecular assisted breeding, ultimately producing novel germplasm that will meet the coming challenges of the 21st Century.

Near-infrared spectroscopy combined with pattern recognition algorithms to quickly classify raisins

With the development of commodity economy, the emergence of fake and shoddy raisin has seriously harmed the interests of consumers and enterprises. To deal with this problem, a classification method combining near-infrared spectroscopy and pattern recognition algorithms were proposed for adulterated raisins. In this study, the experiment was performed by three kinds of raisins in Xinjiang (Hongxiangfei, Manaiti, Munage). After collecting and normalizing the spectral data, we compared the spectra of three kinds of raisins. Next the principal component analysis (PCA) was preformed to compress the dimension of the spectral data, and then classification models including support vector machine (SVM), multiscale fusion convolutional neural network (MCNN) and improved AlexNet were established to identify raisins. The accuracy of SVM, MCNN, and improved AlexNet is 100%, 92.83%, and 97.78% respectively. This study proves that near-infrared spectroscopy combined with pattern recognition is feasible for the raisin inspection.

Near-Infrared Hyperspectral Imaging Pipelines for Pasture Seed Quality Evaluation: An Overview

Near-infrared (800–2500 nm; NIR) spectroscopy coupled to hyperspectral imaging (NIR-HSI) has greatly enhanced its capability and thus widened its application and use across various industries. This non-destructive technique that is sensitive to both physical and chemical attributes of virtually any material can be used for both qualitative and quantitative analyses. This review describes the advancement of NIR to NIR-HSI in agricultural applications with a focus on seed quality features for agronomically important seeds. NIR-HSI seed phenotyping, describing sample sizes used for building high-accuracy calibration and prediction models for full or selected wavelengths of the NIR region, is explored. The molecular interpretation of absorbance bands in the NIR region is difficult; hence, this review offers important NIR absorbance band assignments that have been reported in literature. Opportunities for NIR-HSI seed phenotyping in forage grass seed are described and a step-by-step data-acquisition and analysis pipeline for the determination of seed quality in perennial ryegrass seeds is also presented.

Identification of Rice Seed Varieties Based on Near-Infrared Hyperspectral Imaging Technology Combined with Deep Learning

Rice is one of the most important food crops in the world, and rice seed varieties are related to the yield and quality of rice. This study used near-infrared (NIR) hyperspectral technology with conventional machine learning methods (support vector machine (SVM), logistic regression (LR), and random forest (RF)) and deep learning methods (LeNet, GoogLeNet, and residual network (ResNet)) to establish variety identification models for five common types of rice seeds. Among the deep learning methods, the classification accuracies of most models were higher than 95%. This study further used the deep learning methods to establish variety identification models for 10 varieties of rice seeds without considering their types. Among them, the ResNet model had the best classification results. The classification accuracy on the test set was 86.08%. This study used the saliency map method to visualize each convolutional neural network (CNN) model to find the band region that contributed the most to the data. The results showed that the bands with the largest data contribution were mainly concentrated at approximately 1300-1400 nm and secondarily concentrated at approximately 1050-1250 nm. The overall results showed that NIR hyperspectral imaging technology combined with deep learning could effectively distinguish rice seeds of different varieties. This method provided an effective way to identify rice seed varieties in a quick and nondestructive manner.

Effective identification of varieties by nucleotide polymorphisms and its application for essentially derived variety identification in rice

BACKGROUND

Plant variety identification is the one most important of agricultural systems. Development of DNA marker profiles of released varieties to compare with candidate variety or future variety is required. However, strictly speaking, scientists did not use most existing variety identification techniques for "identification" but for "distinction of a limited number of cultivars," of which generalization ability always not be well estimated. Because many varieties have similar genetic backgrounds, even some essentially derived varieties (EDVs) are involved, which brings difficulties for identification and breeding progress. A fast, accurate variety identification method, which also has good performance on EDV determination, needs to be developed.

RESULTS

In this study, with the strategy of "Divide and Conquer," a variety identification method Conditional Random Selection (CRS) method based on SNP of the whole genome of 3024 rice varieties was developed and be applied in essentially derived variety (EDV) identification of rice. CRS is a fast, efficient, and automated variety identification method. Meanwhile, in practical, with the optimal threshold of identity score searched in this study, the set of SNP (including 390 SNPs) showed optimal performance on EDV and non-EDV identification in two independent testing datasets.

CONCLUSION

This approach first selected a minimal set of SNPs to discriminate non-EDVs in the 3000 Rice Genome Project, then united several simplified SNP sets to improve its generalization ability for EDV and non-EDV identification in testing datasets. The results suggested that the CRS method outperformed traditional feature selection methods. Furthermore, it provides a new way to screen out core SNP loci from the whole genome for DNA fingerprinting of crop varieties and be useful for crop breeding.

Quantitative Analysis of Droplet Size Distribution in Plant Protection Spray Based on Machine Learning Method

Spray droplet size is the main factor affecting the deposition uniformity on a target crop. Studying the influence of multiple factors on the droplet size distribution as well as the evaluation method is of great significance for improving the utilization of pesticides. In this paper, volume median diameter (VMD) and relative span (RS) were selected to evaluate the droplet size distribution under different hollow cone nozzles, flow rates and spatial positions, and the quantitative models of VMD and RS were established based on machine learning methods. The results showed that support vector regression (SVR) had excellent results for VMD (Rc = 0.9974, Rp = 0.9929), while multi-layer perceptron (MLP) had the best effect for RS (Rc = 0.9504, Rp = 0.9537). The correlation coefficient of the prediction set is higher than 0.95, showing the excellent ability of machine learning on predicting the droplet size distribution. In addition, the visualization images of the droplet size distribution were obtained based on the optimal models, which provided intuitive guidance for realizing the uniform distribution of pesticide deposition. In conclusion, this study provides a novel and feasible method for quantitative evaluation of droplet size distribution and offers a theoretical basis for further determining appropriate operation parameters according to the optimal droplet size.

HyperSeed: An End-to-End Method to Process Hyperspectral Images of Seeds

High-throughput, nondestructive, and precise measurement of seeds is critical for the evaluation of seed quality and the improvement of agricultural productions. To this end, we have developed a novel end-to-end platform named HyperSeed to provide hyperspectral information for seeds. As a test case, the hyperspectral images of rice seeds are obtained from a high-performance line-scan image spectrograph covering the spectral range from 600 to 1700 nm. The acquired images are processed via a graphical user interface (GUI)-based open-source software for background removal and seed segmentation. The output is generated in the form of a hyperspectral cube and curve for each seed. In our experiment, we presented the visual results of seed segmentation on different seed species. Moreover, we conducted a classification of seeds raised in heat stress and control environments using both traditional machine learning models and neural network models. The results show that the proposed 3D convolutional neural network (3D CNN) model has the highest accuracy, which is 97.5% in seed-based classification and 94.21% in pixel-based classification, compared to 80.0% in seed-based classification and 85.67% in seed-based classification from the support vector machine (SVM) model. Moreover, our pipeline enables systematic analysis of spectral curves and identification of wavelengths of biological interest.

An Ensemble Deep Learning based Predictor for Simultaneously Identifying Protein Ubiquitylation and SUMOylation Sites

BACKGROUND

Several computational tools for predicting protein Ubiquitylation and SUMOylation sites have been proposed to study their regulatory roles in gene location, gene expression, and genome replication. However, existing methods generally rely on feature engineering, and ignore the natural similarity between the two types of protein translational modification. This study is the first all-in-one deep network to predict protein Ubiquitylation and SUMOylation sites from protein sequences as well as their crosstalk sites simultaneously. Our deep learning architecture integrates several meta classifiers that apply deep neural networks to protein sequence information and physico-chemical properties, which were trained on multi-label classification mode for simultaneously identifying protein Ubiquitylation and SUMOylation as well as their crosstalk sites.

RESULTS

The promising AUCs of our method on Ubiquitylation, SUMOylation and crosstalk sites achieved 0.838, 0.888, and 0.862 respectively on tenfold cross-validation. The corresponding APs reached 0.683, 0.804 and 0.552, which also validated our effectiveness.

CONCLUSIONS

The proposed architecture managed to classify ubiquitylated and SUMOylated lysine residues along with their crosstalk sites, and outperformed other well-known Ubiquitylation and SUMOylation site prediction tools.

Automated In Situ Seed Variety Identification via Deep Learning: A Case Study in Chickpea

On-time seed variety recognition is critical to limit qualitative and quantitative yield loss and asynchronous crop production. The conventional method is a subjective and error-prone process, since it relies on human experts and usually requires accredited seed material. This paper presents a convolutional neural network (CNN) framework for automatic identification of chickpea varieties by using seed images in the visible spectrum (400–700 nm). Two low-cost devices were employed for image acquisition. Lighting and imaging (background, focus, angle, and camera-to-sample distance) conditions were variable. The VGG16 architecture was modified by a global average pooling layer, dense layers, a batch normalization layer, and a dropout layer. Distinguishing the intricate visual features of the diverse chickpea varieties and recognizing them according to these features was conceivable by the obtained model. A five-fold cross-validation was performed to evaluate the uncertainty and predictive efficiency of the CNN model. The modified deep learning model was able to recognize different chickpea seed varieties with an average classification accuracy of over 94%. In addition, the proposed vision-based model was very robust in seed variety identification, and independent of image acquisition device, light environment, and imaging settings. This opens the avenue for the extension into novel applications using mobile phones to acquire and process information in situ. The proposed procedure derives possibilities for deployment in the seed industry and mobile applications for fast and robust automated seed identification practices.

Machine Learning in Agriculture: A Comprehensive Updated Review

The digital transformation of agriculture has evolved various aspects of management into artificial intelligent systems for the sake of making value from the ever-increasing data originated from numerous sources. A subset of artificial intelligence, namely machine learning, has a considerable potential to handle numerous challenges in the establishment of knowledge-based farming systems. The present study aims at shedding light on machine learning in agriculture by thoroughly reviewing the recent scholarly literature based on keywords' combinations of "machine learning" along with "crop management", "water management", "soil management", and "livestock management", and in accordance with PRISMA guidelines. Only journal papers were considered eligible that were published within 2018-2020. The results indicated that this topic pertains to different disciplines that favour convergence research at the international level. Furthermore, crop management was observed to be at the centre of attention. A plethora of machine learning algorithms were used, with those belonging to Artificial Neural Networks being more efficient. In addition, maize and wheat as well as cattle and sheep were the most investigated crops and animals, respectively. Finally, a variety of sensors, attached on satellites and unmanned ground and aerial vehicles, have been utilized as a means of getting reliable input data for the data analyses. It is anticipated that this study will constitute a beneficial guide to all stakeholders towards enhancing awareness of the potential advantages of using machine learning in agriculture and contributing to a more systematic research on this topic.

Advances in infrared spectroscopy combined with artificial neural network for the authentication and traceability of food

The authentication and traceability of food attract more attention due to the increasing consumer awareness regarding nutrition and health, being a new hotspot of food science. Infrared spectroscopy (IRS) combined with shallow neural network has been widely proven to be an effective food analysis technology. As an advanced deep learning technology, deep neural network has also been explored to analyze and solve food-related IRS problems in recent years. The present review begins with brief introductions to IRS and artificial neural network (ANN), including shallow neural network and deep neural network. More notably, it emphasizes the comprehensive overview of the advances of the technology combined IRS with ANN for the authentication and traceability of food, based on relevant literature from 2014 to early 2020. In detail, the types of IRS and ANN, modeling processes, experimental results, and model comparisons in related studies are described to set forth the usage and performance of the combined technology for food analysis. The combined technology shows excellent ability to authenticate food quality and safety, involving chemical components, freshness, microorganisms, damages, toxic substances, and adulteration. As well, it shows excellent performance in the traceability of food variety and origin. The advantages, current limitations, and future trends of the combined technology are further discussed to provide a thoughtful viewpoint on the challenges and expectations of online applications for the authentication and traceability of food.

Classification of Watermelon Seeds Using Morphological Patterns of X-ray Imaging: A Comparison of Conventional Machine Learning and Deep Learning

In this study, conventional machine learning and deep leaning approaches were evaluated using X-ray imaging techniques for investigating the internal parameters (endosperm and air space) of three cultivars of watermelon seed. In the conventional machine learning, six types of image features were extracted after applying different types of image preprocessing, such as image intensity and contrast enhancement, and noise reduction. The sequential forward selection (SFS) method and Fisher objective function were used as the search strategy and feature optimization. Three classifiers were tested (linear discriminant analysis (LDA), quadratic discriminant analysis (QDA), and k-nearest neighbors algorithm (KNN)) to find the best performer. On the other hand, in the transfer learning (deep learning) approaches, simple ConvNet, AlexNet, VGG-19, ResNet-50, and ResNet-101 were used to train the dataset and class prediction of the seed. For the supervised model development (both conventional machine learning and deep learning), the germination test results of the samples were used where the seeds were divided into two classes: (1) normal viable seeds and (2) nonviable and abnormal viable seeds. In the conventional classification, 83.6% accuracy was obtained by LDA using 48 features. ResNet-50 performed better than other transfer learning architectures, with an 87.3% accuracy which was the highest accuracy in all classification models. The findings of this study manifested that transfer learning is a constructive strategy for classifying seeds by analyzing their morphology, where X-ray imaging can be adopted as a potential imaging technique.

Application and interpretation of deep learning methods for the geographical origin identification of Radix Glycyrrhizae using hyperspectral imaging

Radix Glycyrrhizae is used as a functional food and traditional medicine. The geographical origin of Radix Glycyrrhizae is a determinant factor influencing the chemical and physical properties as well as its medicinal and health effects. The visible/near-infrared (Vis/NIR) (376–1044 nm) and near-infrared (NIR) hyperspectral imaging (915–1699 nm) were used to identify the geographical origin of Radix Glycyrrhizae. Convolutional neural network (CNN) and recurrent neural network (RNN) models in deep learning methods were built using extracted spectra, with logistic regression (LR) and support vector machine (SVM) models as comparisons. For both spectral ranges, the deep learning methods, LR and SVM all exhibited good results. The classification accuracy was over 90% for the calibration, validation, and prediction sets by the LR, CNN, and RNN models. Slight differences in classification performances existed between the two spectral ranges. Further, interpretation of the CNN model was conducted to identify the important wavelengths, and the wavelengths with high contribution rates that affected the discriminant analysis were consistent with the spectral differences. Thus, the overall results illustrate that hyperspectral imaging with deep learning methods can be used to identify the geographical origin of Radix Glycyrrhizae, which provides a new basis for related research.

Hyperspectral imaging provides an effective way to identify the geographical origin of Radix Glycyrrhizae to assess its quality.

Wheat Kernel Variety Identification Based on a Large Near-Infrared Spectral Dataset and a Novel Deep Learning-Based Feature Selection Method

Near-infrared (NIR) hyperspectroscopy becomes an emerging nondestructive sensing technology for inspection of crop seeds. A large spectral dataset of more than 140,000 wheat kernels in 30 varieties was prepared for classification. Feature selection is a critical segment in large spectral data analysis. A novel convolutional neural network-based feature selector (CNN-FS) was proposed to screen out deeply target-related spectral channels. A convolutional neural network with attention (CNN-ATT) framework was designed for one-dimension data classification. Popular machine learning models including support vector machine (SVM) and partial least square discrimination analysis were used as the benchmark classifiers. Features selected by conventional feature selection algorithms were considered for comparison. Results showed that the designed CNN-ATT produced a higher performance than the compared classifier. The proposed CNN-FS found a subset of features, which made a better representation of raw dataset than conventional selectors did. The CNN-ATT achieved an accuracy of 93.01% using the full spectra and keep its high precision (90.20%) by training on the 60-channel features obtained via the CNN-FS method. The proposed methods have great potential for handling the analyzing tasks on other large spectral datasets. The proposed feature selection structure can be extended to design other new model-based selectors. The combination of NIR hyperspectroscopic technology and the proposed models has great potential for automatic nondestructive classification of single wheat kernels.

Application of Convolutional Neural Network-Based Feature Extraction and Data Fusion for Geographical Origin Identification of Radix Astragali by Visible/Short-Wave Near-Infrared and Near Infrared Hyperspectral Imaging

Radix Astragali is a prized traditional Chinese functional food that is used for both medicine and food purposes, with various benefits such as immunomodulation, anti-tumor, and anti-oxidation. The geographical origin of Radix Astragali has a significant impact on its quality attributes. Determining the geographical origins of Radix Astragali is essential for quality evaluation. Hyperspectral imaging covering the visible/short-wave near-infrared range (Vis-NIR, 380-1030 nm) and near-infrared range (NIR, 874-1734 nm) were applied to identify Radix Astragali from five different geographical origins. Principal component analysis (PCA) was utilized to form score images to achieve preliminary qualitative identification. PCA and convolutional neural network (CNN) were used for feature extraction. Measurement-level fusion and feature-level fusion were performed on the original spectra at different spectral ranges and the corresponding features. Support vector machine (SVM), logistic regression (LR), and CNN models based on full wavelengths, extracted features, and fusion datasets were established with excellent results; all the models obtained an accuracy of over 98% for different datasets. The results illustrate that hyperspectral imaging combined with CNN and fusion strategy could be an effective method for origin identification of Radix Astragali.

Rapid and Nondestructive Discrimination of Geographical Origins of Longjing Tea using Hyperspectral Imaging at Two Spectral Ranges Coupled with Machine Learning Methods

Longjing tea is one of China’s protected geographical indication products with high commercial and nutritional value. The geographical origin of Longjing tea is an important factor influencing its commercial and nutritional value. Hyperspectral imaging systems covering the two spectral ranges of 380–1030 nm and 874–1734 nm were used to identify a single tea leaf of Longjing tea from six geographical origins. Principal component analysis (PCA) was conducted on hyperspectral images to form PCA score images. Differences among samples from different geographical origins were visually observed from the PCA score images. Support vector machine (SVM) and partial least squares discriminant analysis (PLS-DA) models were built using the full spectra at the two spectral ranges. Decent classification performances were obtained at the two spectral ranges, with the overall classification accuracy of the calibration and prediction sets over 84%. Furthermore, prediction maps for geographical origins identification of Longjing tea were obtained by applying the SVM models on the hyperspectral images. The overall results illustrate that hyperspectral imaging at both spectral ranges can be applied to identify the geographical origin of single tea leaves of Longjing tea. This study provides a new, rapid, and non-destructive alternative for Longjing tea geographical origins identification.

Investigation on Data Fusion of Multisource Spectral Data for Rice Leaf Diseases Identification Using Machine Learning Methods

Rice diseases are major threats to rice yield and quality. Rapid and accurate detection of rice diseases is of great importance for precise disease prevention and treatment. Various spectroscopic techniques have been used to detect plant diseases. To rapidly and accurately detect three different rice diseases [leaf blight (Xanthomonas oryzae pv. Oryzae), rice blast (Pyricularia oryzae), and rice sheath blight (Rhizoctonia solani)], three spectroscopic techniques were applied, including visible/near-infrared hyperspectral imaging (HSI) spectra, mid-infrared spectroscopy (MIR), and laser-induced breakdown spectroscopy (LIBS). Three different levels of data fusion (raw data fusion, feature fusion, and decision fusion) fusing three different types of spectral features were adopted to categorize the diseases of rice. Principal component analysis (PCA) and autoencoder (AE) were used to extract features. Identification models based on each technique and different fusion levels were built using support vector machine (SVM), logistic regression (LR), and convolution neural network (CNN) models. Models based on HSI performed better than those based on MIR and LIBS, with the accuracy over 93% for the test set based on PCA features of HSI spectra. The performance of rice disease identification varied with different levels of fusion. The results showed that feature fusion and decision fusion could enhance identification performance. The overall results illustrated that the three techniques could be used to identify rice diseases, and data fusion strategies have great potential to be used for rice disease detection.

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.

Identification of Soybean Seed Varieties Based on Hyperspectral Imaging Technology

Hyperspectral imaging is a nondestructive testing technology that integrates spectroscopy and iconology technologies, which enables us to quickly obtain both internal and external information of objects and identify crop seed varieties. First, the hyperspectral images of ten soybean seed varieties were collected and the reflectance was obtained. Savitzky-Golay smoothing (SG), first derivative (FD), standard normal variate (SNV), fast Fourier transform (FFT), Hilbert transform (HT), and multiplicative scatter correction (MSC) spectral reflectance pretreatment methods were used. Then, the feature wavelengths and feature information of the pretreated spectral reflectance data were extracted using competitive adaptive reweighted sampling (CARS), the successive projections algorithm (SPA), and principal component analysis (PCA). Finally, 5 classifiers, Bayes, support vector machine (SVM), k-nearest neighbor (KNN), ensemble learning (EL), and artificial neural network (ANN), were used to identify seed varieties. The results showed that MSC-CARS-EL had the highest accuracy among the 90 combinations, with training set, test set, and 5-fold cross-validation accuracies of 100%, 100%, and 99.8%, respectively. Moreover, the contribution of spectral pretreatment to discrimination accuracy was higher than those of feature extraction and classifier selection. Pretreatment methods determined the range of the identification accuracy, feature-selective methods and classifiers only changed within this range. The experimental results provide a good reference for the identification of other crop seed varieties.

Application of Near-Infrared Hyperspectral Imaging with Machine Learning Methods to Identify Geographical Origins of Dry Narrow-Leaved Oleaster (Elaeagnus angustifolia) Fruits

Narrow-leaved oleaster (Elaeagnus angustifolia) fruit is a kind of natural product used as food and traditional medicine. Narrow-leaved oleaster fruits from different geographical origins vary in chemical and physical properties and differ in their nutritional and commercial values. In this study, near-infrared hyperspectral imaging covering the spectral range of 874–1734 nm was used to identify the geographical origins of dry narrow-leaved oleaster fruits with machine learning methods. Average spectra of each single narrow-leaved oleaster fruit were extracted. Second derivative spectra were used to identify effective wavelengths. Partial least squares discriminant analysis (PLS-DA) and support vector machine (SVM) were used to build discriminant models for geographical origin identification using full spectra and effective wavelengths. In addition, deep convolutional neural network (CNN) models were built using full spectra and effective wavelengths. Good classification performances were obtained by these three models using full spectra and effective wavelengths, with classification accuracy of the calibration, validation, and prediction set all over 90%. Models using effective wavelengths obtained close results to models using full spectra. The performances of the PLS-DA, SVM, and CNN models were close. The overall results illustrated that near-infrared hyperspectral imaging coupled with machine learning could be used to trace geographical origins of dry narrow-leaved oleaster fruits.