AI-BASED MOBILE APPLICATION FOR CROP DISEASE RECOGNITION AND MANAGEMENT RECOMMENDATIONS FOR NIGERIAN SMALLHOLDER FARMERS

OLIVER, CHRISTIAN NMESOMA (2026) AI-BASED MOBILE APPLICATION FOR CROP DISEASE RECOGNITION AND MANAGEMENT RECOMMENDATIONS FOR NIGERIAN SMALLHOLDER FARMERS. Other thesis, Godfrey Okoye University, Enugu Nigeria.

[img] Text
CHAPTER ONE TO FIVE .docx....updatedd 200-1.pdf
Download (1MB)

Abstract

In this project, AgroAI Doctor, a mobile-based intelligent system that can detect plant diseases and provide practical guidance to farmers, will be introduced. This paper highlights some drawbacks in current solutions used in agricultural fields, such as reliance on the Internet, no local disease recognition capabilities, and difficulty of use by people with low literacy. The suggested system uses a CNN model, which is based on the MobileNetV2 neural network architecture. Moreover, the optimization of this system is done via TensorFlow Lite for effective model predictions. On the other hand, Flutter is used to develop a mobile app, while FastAPI and Supabase cloud platform are used for implementing an online system and performing online predictions and storage. The user uses the mobile application to take an image of a plant leaf; the image is preprocessed and analyzed either in the cloud or locally based on network availability. Disease classification and localized advice for treatment is provided by the system. It has been shown that AgroAI Doctor works well and delivers accurate predictions even under various conditions. Usability has been considered a priority in implementing AgroAI Doctor using a visual design approach, along with streamlined navigation. In summary, AgroAI Doctor offers an effective solution for farmers in developing countries.

Item Type: Thesis (Other)
Subjects: Q Science > Q Science (General)
Divisions: Faculty of Computing And Information Technology (FACIT)
Depositing User: Cynthia Ugwuoti
Date Deposited: 24 Jul 2026 15:09
Last Modified: 24 Jul 2026 15:09
URI: http://eprints.gouni.edu.ng/id/eprint/6016

Actions (login required)

View Item View Item
GOUNI Repository is powered by EPrints 3 which is developed by the School of Electronics and Computer Science at the University of Southampton. More information and software credits.