UZOCHUKWU, TOCHUKWU BRIGHT (2026) AI-BASED CASSAVA LEAF DISEASE DETECTION AND CLASSIFICATION USING DEEP LEARNING. Other thesis, Godfrey Okoye University, Enugu Nigeria.
|
Text
UZOCHUKWU TOCHUKWU BRIGHT final correction.pdf Download (1MB) | Preview |
Abstract
The reliance on manual, subjective visual inspections for the identification of cassava crop pathologies remains a critical bottleneck in sub-Saharan agricultural productivity, often leading to delayed interventions and significant yield losses. This research presents the design and implementation of a robust, AI-powered predictive system engineered to automate the detection and classification of diseases affecting cassava leaves. Leveraging the YOLOv11 deep learning architecture, the developed system facilitates high-precision, real-time diagnostic capabilities by processing foliar imagery to identify and classify specific pathogens while simultaneously highlighting localized areas of infection. The model was trained and validated using a hybrid dataset—integrating open-source repositories with localized field samples from Nigeria—to ensure superior generalizability across diverse environmental and phenotypic conditions. Adhering to object-oriented design principles, the platform features an intuitive web-based interface that empowers farmers and agricultural extension officers to make data-driven management decisions. Empirical evaluation of the system yielded a Top-1 accuracy of 87.95% and a Top-5 accuracy of 100%, effectively demonstrating the model’s reliability in distinguishing between complex disease patterns. Ultimately, this study provides a scalable, sustainable technological solution to enhance food security and optimize crop health monitoring within regional agricultural workflows.
| Item Type: | Thesis (Other) |
|---|---|
| Subjects: | Q Science > Q Science (General) |
| Divisions: | Faculty of Engineering, Science and Mathematics > School of Electronics and Computer Science |
| Depositing User: | Cynthia Ugwuoti |
| Date Deposited: | 22 Jul 2026 12:45 |
| Last Modified: | 22 Jul 2026 13:18 |
| URI: | http://eprints.gouni.edu.ng/id/eprint/6017 |
Actions (login required)
 |
View Item |
