Tomato Leaf Disease Detection Using Transfer Learning: A Comparative Study

Tomato, a vital food crop enriched with essential minerals and vitamins, is widely used in various cuisines. Leaf disease is a common problem that significantly impacts the quantity and quality of tomato production. So, farmers should accurately surveil tomato farms to manufacture high-quality foods...

Full description

Saved in:
Bibliographic Details
Main Authors: Bahrami, Mahsa, Pourhatami, Aliakbar, Maboodi, Mohsen
Format: Conference Proceeding
Language:English
Subjects:
Deep learning
detection
Feature extraction
leaf disease
Pesticides
Plant diseases
Production
Smart agriculture
tomato
Transfer learning
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Tomato, a vital food crop enriched with essential minerals and vitamins, is widely used in various cuisines. Leaf disease is a common problem that significantly impacts the quantity and quality of tomato production. So, farmers should accurately surveil tomato farms to manufacture high-quality foods free from pests. In this regard, accurate and robust algorithms are needed to detect tomato leaf disease. Among various methods, deep learning algorithms provide automatic, accurate, and robust leaf disease detection algorithms. In this paper, a comprehensive analysis of different transfer learning algorithms, including VGG19, ResNet-101, and MobileNet-v2 was done on PlantVillage and CCMT datasets. The best performance was achieved by VGG19, where the best accuracy, precision, recall, and F1-score on the test set of PlantVillage and CCMT datasets were 99.48%, 99.27%, 99.28%, 99.27%, and 92.76%, 92.74%, 95.09%, 90.86%, respectively. The results demonstrate that VGG19 can detect tomato leaf disease precisely and robustly.
ISSN:2166-6784
DOI:10.1109/MVIP62238.2024.10491178