Clinical auxiliary decision-making is related to life and health of patients, so the deep model needs to extract the personalised representation of patients to ensure high analysis and prediction accuracy;and provide ...
详细信息
Detection of orange ripeness is an important part of modern precision agriculture to achieve optimum harvesting, decreased postharvest losses and better orange quality. In this paper, based on the feature extraction a...
详细信息
ISBN:
(纸本)9798331542375
Detection of orange ripeness is an important part of modern precision agriculture to achieve optimum harvesting, decreased postharvest losses and better orange quality. In this paper, based on the feature extraction ability of Convolutional Neural Networks (CNNs) and superior classification power of Support Vector Machines (SVMs), we present a hybrid Convolutional Neural Network - Support Vector Machine (CNN - SVM) framework for automated orange ripeness classification. The proposed model was trained and tested on a custom dataset of 5,000 labeled orange images, categorized into three ripeness stages: unripe, ripe, and overripe. Experimental results demonstrated that the CNN-SVM framework achieved an impressive classification accuracy of 97.5%, significantly outperforming standalone CNN (accuracy: 95. Our proposed approach achieved an accuracy of 85.6% and outperformed baseline eyewitness based models (accuracy: 76.2%), Support Vector Machine Convolutional Neural Networks (accuracy: 81.8%) and traditional SVM models (accuracy: 85.4%). The CNN-SVM model achieved precision and recall values of 96.7% and 97.3%, respectively, for distinguishing ripeness categories with minimum misclassifications, and comparison of the model with alternative methods like LSTM, Federated Learning CNN, CNN-RF and, RF validated its effectiveness in classifying ripeness categories. Our proposed approach also displayed a high area under the curve (AUC) value of 0.98, indicating potent discriminatory power. The analysis of confusion matrix shows very minimal misclassification and occurs between adjacent ripeness stages. Findings show the CNN-SVM framework represents an efficient, scalable, and accurate method for real time ripeness detection and actionable insights for the farmer to optimize harvesting decisions. As the future research, the model will be deployed to real world scenarios through IoT based systems in order to improve smart agricultural practices and increase productivity in citr
In the thickness measurement of reconstituted tobacco by modulated line laser, it is necessary to extract the edge of the modulated line laser to perform the straight line fitting. However, in the edge detection, the ...
详细信息
Wireless Local Area Network (WLAN) Fingerprinting-based Indoor Positioning Systems (IPS) offer several key advantages over other indoor positioning technologies, including cost-effectiveness with high accuracy and pas...
详细信息
Sod culture (SC) and conventional agriculture (CA) represent two distinct field management approaches utilized in the cultivation of tea plants in Taiwan. In this study, we employed gas exchange and chlorophyll fluore...
详细信息
Sulfide-based solid-state lithium-ion batteries (SSLIB) have attracted a lot of interest globally in the past few years for their high safety and high energy density over the traditional lithium-ion batteries. However...
详细信息
This paper proposes a CNN-GRU and dual attention mechanism (DAM) hybrid neural network model (CNN-GRU-DAM) for short-time power load forecasting. First, historical data is transformed into a time-series form, and time...
详细信息
The increasing complexity and interconnectedness of Internet of Things (IoT) software systems necessitate the development of intelligent solutions for predictive maintenance and security. Conventional techniques often...
详细信息
In the working process of the upper stage integrated navigation information fusion system, the multi-source navigation information fusion algorithm based on factor graph Bayesian estimation is used to fuse the informa...
详细信息
Clothing design, as a comprehensive art with commercial value, is inherently connected to abstract painting art, with the two mutually influencing each other. Therefore, fashion designers can use abstract elements to ...
详细信息
暂无评论