Plant Disease Detection Using Machine Learning in a Serverless Environment

Yu, Peng (2024) Plant Disease Detection Using Machine Learning in a Serverless Environment. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (5MB) | Preview
Preview	PDF (Configuration Manual) Download (1MB) | Preview

Abstract

Plant disease is a significant threat to food security, the crop loss caused by pests and diseases is up to 40% according to the United Nations. Compared to on-site checking by agricultural experts, Artificial Intelligence (AI) technologies especially Convolutional Neural Networks (CNN) have a great performance to detect plant disease. However, the seasonal and fluctuating characteristics of agricultural production especially for family farms require a more flexible, scalable, cost-effective solution. In this study, we proposed a plant disease detection system using Deep Learning (DL) in a serverless environment, which potentially helps smallholders recognize crop disease accurately and maintain an affordable budget. We trained and compared three models on top of three CNNs including VGG19, MobileNet-V2 and ResNet50. The evaluation result shows that the MobileNet-V2 has the best performance to detect diseases on vegetables, reaching a 98.82% accuracy. Then we compared the latency and cost by deploying fine-tuned MobileNet-V2 on AWS EC2 and AWS Lambda. The experiment allows us to define a threshold to estimate the type of service that reduces cost according to the number of requests and confirmed that the annual cost of Lambda is 88% lower than that of EC2 within the acceptable range of maintaining latency for the fluctuating agricultural requirement. Finally, our system has a mobile application, a MobileNet-V2 model deployed on the Lambda, providing accuracy, flexible, scalable, and low cost plant disease detection service for smallholder farmers. Theoretically, this system will contribute to food security, especially family farms that account for 90% of all farms globally

Item Type:	Thesis (Masters)
Supervisors:	Name Email Cortes Mendoza, Jorge Mario UNSPECIFIED
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 > QH Natural history > QH301 Biology > Methods of research. Technique. Experimental biology > Data processing. Bioinformatics > Artificial intelligence Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Artificial intelligence T Technology > T Technology (General) > Information Technology > Cloud computing Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning
Divisions:	School of Computing > Master of Science in Cloud Computing
Depositing User:	Ciara O'Brien
Date Deposited:	17 Jul 2025 13:54
Last Modified:	17 Jul 2025 13:54
URI:	https://norma.ncirl.ie/id/eprint/8168

Actions (login required)

View Item