•  
  •  
 

Abstract

Plant diseases pose significant challenges to agricultural productivity and food security, particularly in regions like Oman, where citrus crops, such as lemon trees, are economically vital. Traditional diagnosis methods are time-consuming and often require expert intervention, limiting accessibility for small-scale farmers. This paper presents an AI-driven plant health assessment Nuwriq application. It leverages both image-based and text-based input to diagnose lemon plant diseases, particularly the Lemon Witch Broom Disease (LWBD). It incorporates two machine learning models: a convolutional neural network (CNN) for image classification and a natural language processing (NLP) model for symptom-based textual analysis. Users can upload images of diseased plants or enter descriptive text to receive a diagnosis and recommended remediation. The CNN model achieved higher performance across accuracy and precision metrics compared to the NLP model. Designed with accessibility and real-time feedback in mind, Nuwriq empowers local farmers and agricultural professionals to make informed decisions quickly. The paper aligns with national goals such as Oman Vision 2040 by promoting technological innovation in agriculture and enhancing sustainability. Future improvements include multilingual support, expanded coverage of crops and diseases, and integration with expert feedback.

Download

Share

COinS