Assessing the potential of Convolutional Neutral Network (CNN) in the detection of diseases of cucurbits
| dc.contributor.author | Vashtine P. Liwanagan and Imman Dieve R. Yarcia | |
| dc.date.accessioned | 2026-03-10T00:21:31Z | |
| dc.date.available | 2026-03-10T00:21:31Z | |
| dc.date.issued | 2024 | |
| dc.description.abstract | Cucurbits, including cucumbers, melons, pumpkins, squash and gourds, are part of Cucurbitaceae family with a millennia-long historical significance. They have been pivotal to civilization globally, contributing to food security, economic growth, and cultural heritage. In modern agriculture, cucurbits remain indispensable, thriving in diverse climates, providing nutrition and promoting sustainable farming practices. This study utilizes computer vision and machine learning techniques to tackle challenges in cucurbit production, explicitly focusing on disease detection. The objective is to harness convolutional neutral networks CNN) for early disease identification in cucurbit crops, progressing through three phases: identifying common diseases, developing a CNN-based smart detection system, and evaluating system accuracy and reliability. The methodology involves designing a conceptual framework to guide CNN model training for disease detection. Sample images collected from agricultural fields are standardized using an Image data generator. The CNN model is trained to extract relevant features from images, enabling disease classification. Evaluation metrics, including precision, recall, and F1-score, are analyzed using a confusion matrix. Results show promising performance, with an overall accuracy of 73%. Additionally, the study demonstrates the model's improved performance with increasing training epochs, indicating its real-world application potential. In conclusion, this research underscores the significance of computer vision in revolutionizing disease management strategies in cucurbit production. By integrating technology into agricultural practices, farmers can enhance disease detection, mitigate yield losses, and ensure the sustainability and productivity of cucurbit crops. | |
| dc.identifier.uri | http://granarium.clsu.edu.ph/handle/123456789/1011 | |
| dc.language.iso | en_US | |
| dc.relation.supervisor | NICASIO C. SALVADOR, M.Sc. | |
| dc.title | Assessing the potential of Convolutional Neutral Network (CNN) in the detection of diseases of cucurbits | |
| dc.type | Thesis |