Disease detection of solanaceous crops using deep learning method

A. Halim, Nurul Hidayah (2022) Disease detection of solanaceous crops using deep learning method. Project Report. Universiti Teknikal Malaysia Melaka, Melaka, Malaysia. (Submitted)

	Text (24 Pages) Disease detection of solanaceous crops using deep learning method.pdf - Submitted Version Download (389kB)
	Text (Full text) Disease detection of solanaceous crops using deep learning method.pdf - Submitted Version Restricted to Repository staff only Download (2MB)

Abstract

Nowadays, Artificial Intelligence has become a synonym in our lives. The technology of AI has helped so many people to solve their problems. In agriculture, crop diseases are more likely to occur because of many factors such as shifting weather, lack of nutrition, and pest attacks. This study looks into the solanaceous crops disease for leaf and its fruit. Four types of solanaceous crops were used to be investigated in this project, namely tomato, pepper, eggplant, and potato. The deep learning method of the CNN architecture-based YOLOv5 model was applied in this project to detect the disease of solanaceous crops. The work involved four types of crops, including 23 classes of healthy and disease crops that infected the leaf and fruits. The total dataset of all 23 classes is 16580 images and has been divided into three parts: training set, validation set, and testing set. The dataset that was used for training is 88% of the total dataset (15000 images), 8% of the dataset performed a validation process (1400 images), and the rest of the 4% dataset was used for the test process (699 images). The performance of YOLOv5 has been compared with YOLOv4 to prove that this model is more robust and better in terms of accuracy. The training, validation, and testing dataset simulation was performed on a Google Colab notebook.

Actions (login required)

View Item