|
Applying Convolutional Neural Networks for Sugarcane Leaf Disease Image Classification and Integration with LINE Chatbot to Support Smart Agriculture |
|---|---|
| รหัสดีโอไอ | |
| Creator | Supachai Baipod |
| Title | Applying Convolutional Neural Networks for Sugarcane Leaf Disease Image Classification and Integration with LINE Chatbot to Support Smart Agriculture |
| Contributor | Charuay Savithi, Jackaphan Sriwongsa |
| Publisher | Department of Computer Education, Faculty of Science and Technology, Surindra Rajabhat University |
| Publication Year | 2568 |
| Journal Title | Journal of Computer and Creative Technology |
| Journal Vol. | 3 |
| Journal No. | 2 |
| Page no. | 317-332 |
| Keyword | Convolutional Neural Networks, Leaf Disease Diagnosis, Image Classification, Line Chatbot, Smart Agriculture |
| URL Website | https://so13.tci-thaijo.org/index.php/jcct |
| Website title | Journal of Computer and Creative Technology |
| ISSN | ISSN 2985-1580 (Print);ISSN 2985-1599 (Online) |
| Abstract | Plant leaf diseases, particularly sugarcane leaf diseases in Thailand, represent a significant problem affecting sugarcane yield and quality. Traditional disease detection methods are often time-consuming and require specialized expertise. Therefore, this research aims to develop a deep learning model for sugarcane leaf disease diagnosis by applying Convolutional Neural Networks (CNN) techniques and integrating them with LINE chatbot to support smart agriculture. This study compared the performance of five CNN architectures to identify the optimal model for sugarcane leaf disease classification, including VGGNet-16, ResNet-50, DenseNet-121, AlexNet, and GoogLeNet, using the Sugarcane Leaf Disease Dataset collected by the researchers from real-world environments. The sugarcane leaf image dataset was collected from four provinces in the central northeastern region, totaling 4,000 images divided into four categories of 1,000 images each, covering four types of sugarcane leaves: red stripe disease, ring spot disease, rust disease, and healthy sugarcane leaves. The data was split into three ratios of 75:25, 80:20, and 90:10 for training and testing sets, respectively. Experimental results showed that the DenseNet-121 model with a 90:10 ratio achieved the highest performance, with a Macro-average of 97.28%. When the best model was developed into a LINE Chatbot system for automated sugarcane leaf disease diagnosis, system testing demonstrated effective user response capabilities. This research demonstrates the potential of deep learning technology to support farmers in quickly and accurately monitoring sugarcane plant health, which is crucial for developing smart agriculture systems in Thailand. |
