Penerapan Arsitektur EfficientNet Untuk Klasifikasi Penyakit Daun Jeruk Siam Menggunakan Metode Convolutional Neural Network

Acarya, Burhan Syarif (2024) Penerapan Arsitektur EfficientNet Untuk Klasifikasi Penyakit Daun Jeruk Siam Menggunakan Metode Convolutional Neural Network. Undergraduate thesis, UPN Veteran Jawa Timur.

	Text (Cover) 20083010004 -cover.pdf Download (863kB)
	Text (Bab 1) 20083010004 -bab1.pdf Download (48kB)
	Text (Bab 2) 20083010004 -bab2.pdf Restricted to Repository staff only until 30 May 2026. Download (704kB)
	Text (Bab 3) 20083010004 -bab3.pdf Restricted to Repository staff only until 30 May 2026. Download (527kB)
	Text (Bab 4) 20083010004 -bab4.pdf Restricted to Repository staff only until 30 May 2026. Download (995kB)
	Text (Bab 5) 20083010004 -bab5.pdf Download (31kB)
	Text (Daftar Pustaka) 20083010004 -daftarpustaka.pdf Download (166kB)
	Text (Lampiran) 20083010004 -lampiran.pdf Restricted to Repository staff only Download (898kB)

Abstract

The issue of diseases in citrus plants is crucial in agriculture because it can significantly affect the productivity and quality of citrus fruits. Farmers often lack the knowledge to identify and diagnose diseases in Siam citrus leaves, which frequently leads to errors. This research aims to find a solution by detecting diseases in Siam citrus leaves using the Convolutional Neural Network (CNN) method with EfficientNet architecture. Data of diseased and healthy citrus leaves were collected from an orchard in Semboro District, Jember Regency, East Java. The classes used are Healthy Leaf, Greening Leaf, Canker Leaf, Citrus Leafminer, Blackspot Leaf, and Powdery Mildew. Experiments were conducted using CNN models with EfficientNet-B4 (pretrained & scratch), DenseNet-121 (pretrained & scratch), and ResNet-50 (pretrained & scratch) architectures. Each model was tested with various optimizers (Adam, SGD, and RMSprop), learning rates (0.1, 0.01, 0.001, 0.0001, and 0.00001), and 50 epochs. The research results showed that the EfficientNet architecture with RMSprop optimizer and a learning rate of 0.0001 achieved the highest accuracy: 97% for training data and 90% for testing data. These findings indicate that EfficientNet is effective in detecting diseases in Siam citrus leaves, offering promising support for sustainable agricultural practices through technology.

Item Type:

Thesis (Undergraduate)

Contributors:

Contribution	Contributors	NIDN/NIDK	Email
Thesis advisor	Muhaimin, Amri	0023079502	amri.muhaimin.stat@upnjatim.ac.id
Thesis advisor	Hindrayani, Kartika Maulida	0009099205	kartika.maulida.ds@upnjatim.ac.id

Subjects:

Q Science > QA Mathematics > QA76.6 Computer Programming

Divisions:

Faculty of Computer Science > Departemen of Data Science

Depositing User:

Burhan Syarif Acarya

Date Deposited:

30 May 2024 08:21

Last Modified:

30 May 2024 08:21

URI:

https://repository.upnjatim.ac.id/id/eprint/23551