Detection of the moldy status of the stored maize kernels using hyperspectral imaging and deep learning algorithms

Research on nondestructive detection of sweet-waxy corn seed varieties and mildew based on stacked ensemble learning and hyperspectral feature fusion technology

The variety and quality of corn seeds are crucial factors affecting crop yield and farmers' economic benefits. This study adopts an innovative method based on a hyperspectral imaging system combined with stacked ensemble learning, aiming to achieve varieties classification and mildew detection of sweet-waxy corn seeds. First, data interference is eliminated by extracting the spectral and texture information of each corn sample and preprocessing the data. Secondly, a stacked ensemble learning model (Stack) was constructed by stacking base models and meta-models. Its results were compared with those of the base models, including Gradient Boosting Decision Tree (GBDT), Extreme Gradient Boosting (XGBoost), Light Gradient Boosting Machine (LightGBM), and Random Forest (RF).Finally, the overall performance of the model is improved through the information fusion strategy of hyperspectral data and texture information. The research results indicate that the GBDT-Stack model, which integrates spectral and texture data, demonstrated optimal performance in the comprehensive classification of both corn seed varieties and mold detection. On the test set, the model achieved an average prediction accuracy of 97.01%. Specifically, the model achieved a test set accuracy ranging from 94.49% to 97.58% for different corn seed varieties and a test set accuracy of 98.89% for mildew detection. This model not only classifies corn seed varieties but also accurately detects mildew, demonstrating its wide applicability. The method has huge potential and is of great significance for improving crop yield and quality.

Digital techniques and trends for seed phenotyping using optical sensors

BACKGROUND

The breeding of high-quality, high-yield, and disease-resistant varieties is closely related to food security. The investigation of breeding results relies on the evaluation of seed phenotype, which is a key step in the process of breeding. In the global digitalization trend, digital technology based on optical sensors can perform the digitization of seed phenotype in a non-contact, high throughput way, thus significantly improving breeding efficiency.

AIM OF REVIEW

This paper provides a comprehensive overview of the principles, characteristics, data processing methods, and bottlenecks associated with three digital technique types based on optical sensors: spectroscopy, digital imaging, and three-dimensional (3D) reconstruction techniques. In addition, the applicability and adaptability of digital techniques based on the optical sensors of maize seed phenotype traits, namely external visible phenotype (EVP) and internal invisible phenotype (IIP), are investigated. Furthermore, trends in future equipment, platform, phenotype data, and processing algorithms are discussed. This review offers conceptual and practical support for seed phenotype digitization based on optical sensors, which will provide reference and guidance for future research.

KEY SCIENTIFIC CONCEPTS OF REVIEW

The digital techniques based on optical sensors can perform non-contact and high-throughput seed phenotype evaluation. Due to the distinct characteristics of optical sensors, matching suitable digital techniques according to seed phenotype traits can greatly reduce resource loss, and promote the efficiency of seed evaluation as well as breeding decision-making. Future research in phenotype equipment and platform, phenotype data, and processing algorithms will make digital techniques better meet the demands of seed phenotype evaluation, and promote automatic, integrated, and intelligent evaluation of seed phenotype, further helping to lessen the gap between digital techniques and seed phenotyping.

Non-destructive detection of defective maize kernels using hyperspectral imaging and convolutional neural network with attention module

Rapid, effective and non-destructive detection of the defective maize kernels is crucial for their high-quality storage in granary. Hyperspectral imaging (HSI) coupled with convolutional neural network (CNN) based on spectral and spatial attention (Spl-Spal-At) module was proposed for identifying the different types of maize kernels. The HSI data within 380-1000 nm of six classes of sprouted, heat-damaged, insect-damaged, moldy, broken and healthy kernels was collected. The CNN-Spl-At, CNN-Spal-At and CNN-Spl-Spal-At models were established based on the spectra, images and their fusion features as inputs for the recognition of different kernels. Further compared the performances of proposed models and conventional models were built by support vector machine (SVM) and extreme learning machine (ELM). The results indicated that the recognition ability of CNN with attention series models was significantly better than that of SVM and ELM models and fused features were more conducive to expressing the appearance of different kernels than single features. And the CNN-Spl-Spal-At model had an optimal recognition result with high average classification accuracy of 98.04 % and 94.56 % for the training and testing sets, respectively. The recognition results were visually presented on the surface image of kernels with different colors. The CNN-Spl-Spal-At model was built in this study could effectively detect defective maize kernels, and it also had great potential to provide the analysis approaches for the development of non-destructive testing equipment based on HSI technique for maize quality.

A multivariate algorithm for identifying contaminated peanut using visible and near-infrared hyperspectral imaging

In this study, a novel uniform manifold approximation and projection combined-improved simultaneous optimization genetic algorithm-convolutional neural network (UMAP-ISOGA-CNN) algorithm was proposed. The improved simultaneous optimization genetic algorithm (ISOGA) combined with convolutional neural network (CNN) to optimize the architecture, hyperparameters, and optimizer of the CNN model simultaneously. Additionally, a uniform manifold approximation and projection (UMAP) method was used to visualize the feature space of all feature layers of the ISOGA-CNN model. The UMAP-ISOGA-CNN algorithm combined with visible and near-infrared hyperspectral imaging was used to identify peanut kernels contaminated with Aspergillus flavus and to distinguish their storage time, which is essential for the food industry to monitor the freshness of products. Overall, the UMAP-ISOGA-CNN algorithm provides useful insights into the feature space of the ISOGA-CNN model, contributing to a better understanding of the model's internal mechanisms. This study has practical implications for the food industry and future research on deep learning optimization.

Intelligent grading method for walnut kernels based on deep learning and physiological indicators

Walnut grading is an important step before the product enters the market. However, traditional walnut grading primarily relies on manual assessment of physiological features, which is difficult to implement efficiently. Furthermore, walnut kernel grading is, at present, relatively unsophisticated. Therefore, this study proposes a novel deep-learning model based on a spatial attention mechanism and SE-network structure to grade walnut kernels using machine vision to ensure accuracy and improve assessment efficiency. In this experiment, we found through the literature that both the lightness (L* value) and malondialdehyde (MDA) contens of walnut kernels were correlated with the oxidation phenomenon in walnuts. Subsequently, we clustered four partitionings using the L* values. We then used the MDA values to verify the rationality of these partitionings. Finally, four network models were used for comparison and training: VGG19, EfficientNetB7, ResNet152V2, and spatial attention and spatial enhancement network combined with ResNet152V2 (ResNet152V2-SA-SE). We found that the ResNet152V2-SA-SE model exhibited the best performance, with a maximum test set accuracy of 92.2%. The test set accuracy was improved by 6.2, 63.2, and 74.1% compared with that of ResNet152V2, EfficientNetB7, and VGG19, respectively. Our testing demonstrated that combining spatial attention and spatial enhancement methods improved the recognition of target locations and intrinsic information, while decreasing the attention given to non-target regions. Experiments have demonstrated that combining spatial attention mechanisms with SE networks increases focus on recognizing target locations and intrinsic information, while decreasing focus on non-target regions. Finally, by comparing different learning rates, regularization methods, and batch sizes of the model, we found that the training performance of the model was optimal with a learning rate of 0.001, a batch size of 128, and no regularization methods. In conclusion, this study demonstrated that the ResNet152V2-SA-SE network model was effective in the detection and evaluation of the walnut kernels.