NORMA eResearch @NCI Library

An Implementation of Transfer Learning & Deep Learning Techniques to Detect Tomato Leaf Diseases

Gudivada, Manikanta Dinesh (2020) An Implementation of Transfer Learning & Deep Learning Techniques to Detect Tomato Leaf Diseases. Masters thesis, Dublin, National College of Ireland.

[thumbnail of Master of Science]
Preview
PDF (Master of Science)
Download (843kB) | Preview
[thumbnail of Configuration manual]
Preview
PDF (Configuration manual)
Download (1MB) | Preview

Abstract

From the past few decades, Agriculture is playing an indispensable role in the survival of humans. In agriculture, the most cultivated crops are potato and sweet potato. According to the statistics of FAO organization, the India ranks 3 for their immense production in tomato (Kaur and Bhatia, 2019). Across the world, the tomato crop is enormously popular for their growth as in every kitchen people use tomatoes massively. However, the growth is high there is a huge loss to the farmers due to an increase in massive Tomato Leaf Diseases. So, to detect leaf diseases the majorly used techniques in this paper are Deep Learning techniques where the image data can be trained and modelled perfectly using various Neural Network models. By this research study, every individual can gain knowledge over different methodologies involved in Deep Learning, to encounter those methodologies, this research had experimented 4 models which are Le Net, Dense Net-121, Mobile Net and CNN. Once after building the model to verify the predicted results and this research consists of different evaluation metrics such as Confusion Matrix, Accuracy, Precision, MSE, Recall and F1- Score. To develop, all these models the coding implementation was done in python using Anaconda Navigator and Google Collab software’s. Finally, all the 4 models have gained very good accuracy which is around 85-95%. Out of all the models, Dense-Net model had gained the best accuracy compared to all the models with 97%. All the other models' Le Net, Mobile Net and CNN have gained 86%, 88% and 88% respectively with 50 Epochs. This research had also verified by using 25 and 100 Epochs as well, but Dense Net have given the best results.
Keywords: Deep Learning, Transfer learning, Leaf Disease Detection, Le-Net, Mobile-Net, DenseNet-121, Convolutional Neural Network

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
Divisions: School of Computing > Master of Science in Data Analytics
Depositing User: Dan English
Date Deposited: 20 Jan 2021 14:17
Last Modified: 20 Jan 2021 14:17
URI: https://norma.ncirl.ie/id/eprint/4395

Actions (login required)

View Item View Item