NORMA eResearch @NCI Library

Machine Learning Based Approach in Detection and Classification of Tomato Plant Leaf Diseases

Ramakrishna, Rajath (2020) Machine Learning Based Approach in Detection and Classification of Tomato Plant Leaf Diseases. Masters thesis, Dublin, National College of Ireland.

[thumbnail of Master of Science]
PDF (Master of Science)
Download (2MB) | Preview
[thumbnail of Configuration manual]
PDF (Configuration manual)
Download (1MB) | Preview


Tomatoes are one of the most commonly and widely grown crop across the world. As they are one of the staple plant-produces, they are extensively used in every commercial and household kitchen. It is an integral part of the human diet in every continent and culture. Hence tomato cultivation is of great interest to agriculturists. However, tomato plants are not immune to plant diseases. These plant diseases easily and immediately affect the quantity of produce as well as the quality of produce. Therefore, monitoring and analyzing the growth stages of the crop can promote towards the production of disease-free produce with minimal losses to the crop. This research is an attempt at contributing towards the early detection and identification of the onset of diseases in tomato leaves using machine learning. This study intends to analyze the efficiency of algorithms such as XGBoost, Convolutional Neural Network (CNN) and its architectures such as VGG16 and VGG19 in combination with data augmentation and transfer learning over conventional machine learning algorithms such as Support Vector Machines (SVM), Random Forest and test their efficiency in the detection and classification of tomato plant leaf diseases in terms of accuracy, precision, recall and training time. Models training is conducted on publicly accessible PlantVillage dataset consisting of approximately 6500 images of both healthy and diseased tomato plant leaves encompassing five distinct classes. CNN's VGG19 architecture with transfer learning achieved the best result with the overall accuracy of 96 percent compared to other models while having faster training time. Also, it is observed that the performance of the models built improves with increase in the training data. The results achieved on the architectures of CNN while implemented along the transfer learning yields promising results therefore it can be explored more in depth to create an efficient automated leaf disease diagnosis model and thus help farmers and other researchers in agricultural community in identifying and classifying tomato plant leaf diseases.

Item Type: Thesis (Masters)
Subjects: Q Science > QA Mathematics > Electronic computers. Computer science
T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science
Q Science > QA Mathematics > Computer software
T Technology > T Technology (General) > Information Technology > Computer software
S Agriculture > S Agriculture (General) > Farming Industry
Divisions: School of Computing > Master of Science in Data Analytics
Depositing User: Dan English
Date Deposited: 11 Jun 2020 12:46
Last Modified: 11 Jun 2020 12:46

Actions (login required)

View Item View Item