Efficacy of Deep Learning Model for Plant Disease Classification With Limited Data

Sharma, Saurabh (2022) Efficacy of Deep Learning Model for Plant Disease Classification With Limited Data. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (1MB) | Preview
Preview	PDF (Configuration manual) Download (1MB) | Preview

Abstract

Plants, like humans, are prone to a wide range of diseases, which have ramifications not just for human health but also for the country’s economic development. Hence, it is essential to detect plant disease early to avoid such unfavourable effects. As per the research done in this field, deep learning models work well in classifying plant disease. However, it has a drawback: it needs a large amount of data to get trained. Plant leaf data collection is a tedious task and requires manual effort. This paper focuses on validating the efficacy of the deep learning model with a small amount of data set for plant classification. The plant disease dataset has three types of leaves: healthy, powdery and rust. It is seen from the research that a pre-trained MobileNetV2 model performs well with this limitation. The research uses ResNet50, a pre-trained MobileNetv2 and a hybrid CNN-RF for plant disease classification. Out of the three models, MobileNetV2 achieves an accuracy of over 95% and has a precision and recall value of 0.97. The model is also predicting unseen images correctly. The MobileNetV2 model is then converted as a TensorFlow Lite model, ready to be deployed in mobile for real-life prediction.

Item Type:	Thesis (Masters)
Subjects:	Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science S Agriculture > SB Plant culture Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning
Divisions:	School of Computing > Master of Science in Data Analytics
Depositing User:	Tamara Malone
Date Deposited:	10 Mar 2023 17:57
Last Modified:	10 Mar 2023 17:57
URI:	https://norma.ncirl.ie/id/eprint/6301

Actions (login required)

View Item