Abstract:
Efficient along with precise identification of plant diseases can significantly enhance
agricultural productivity and overall crop yield, particularly for lemon trees and their
foliage.[24] Algal Leaf Spot, Black spot, Citrus canker, Citrus pest, Citrus Scab, Dry leaf,
Greening, Leaf Curl, and yellow spot are diseases that can significantly impair the health
of lemon trees and reduce their fruit yield. These diseases typically develop on the leaves,
leading to reduced fruit quality and output, sometimes by more than half. Convolutional Neural Networks (CNN) have demonstrated promise in the field of lemon
tree disease identification.[25] Nevertheless, there is a lack of research specifically focused
on the identification of diseases in lemon trees, particularly in regions such as Bangladesh
where lemon cultivation is widespread. Previous experiments in this field have primarily
based on secondary data, limiting the depth of knowledge and practical application. In order to fill these knowledge gaps, our research focused on collecting lemon disease
leaves from well-known lemon-growing regions, such as Khagan and Ashulia in
Bangladesh. I utilized advanced Convolutional Neural Network (CNN) techniques,
specifically YOLOv8, to assess its effectiveness in identifying diseases. The trials I
conducted highlighted the benefits of YOLOv8 in terms of its speed and accuracy,
demonstrating its practicality for identifying diseases in agricultural settings. Furthermore, our review of current literature uncovered a variety of machine-learning
methods applied to publicly available datasets for lemon leaf disease diagnosis. By doing
our experiments, I obtained a precision confidence curve score of 1.00, indicating a high
level of confidence in the accuracy of our data. Various metrics, including precision,
accuracy, and recall, were utilized to examine the performance of the model.
