DETEKSI TUBERKULOSIS PADA CITRA X - RAY MENGGUNAKAN METODE HISTOGRAM OF ORIENTED GRADIENTS DAN BACKPROPAGATION NEURAL NETWORK

Puspitaningrum, Ananda Ayu (2025) DETEKSI TUBERKULOSIS PADA CITRA X - RAY MENGGUNAKAN METODE HISTOGRAM OF ORIENTED GRADIENTS DAN BACKPROPAGATION NEURAL NETWORK. Undergraduate thesis, UPN Veteran Jawa Timur.

	Text (Cover) 21081010242-cover.pdf Download (1MB)
	Text (Bab 1) 21081010242-bab1.pdf Download (145kB)
	Text (Bab 2) 21081010242-bab2.pdf Restricted to Repository staff only until 5 December 2027. Download (520kB)
	Text (Bab 3) 21081010242-bab3.pdf Restricted to Repository staff only until 5 December 2027. Download (482kB)
	Text (Bab 4) 21081010242-bab4.pdf Restricted to Repository staff only until 5 December 2027. Download (1MB)
	Text (Bab 5) 21081010242-bab5.pdf Download (135kB)
	Text (Daftar Pustaka) 21081010242-daftarpustaka.pdf Download (82kB)
	Text (Lampiran) 21081010242-lampiran.pdf Restricted to Repository staff only Download (211kB)

Abstract

Tuberculosis is an infectious disease that attacks the lungs and can be detected through chest X – ray images. This study aims to develop a tuberculosis detection system on X – ray images using the Histogram of Oriented Gradients (HOG) method as feature extraction and Backpropagation Neural Network (BPNN) as a classification model. The research data consists of 7.000 X – ray images divided into two classes, namely normal and tuberculosis. The research stages included preprocessing (grayscale, resize to 256 x 256 pixels, histogram equalization, and median filter), followed by HOG feature extraction to capture the shape and edge patterns and gradient direction with cell size parameters of 8 x 8 pixels, 2 x 2 cell blocks, and 9 orientation bins. The BPNN model was trained with variations in learning rate, epoch, number of hidden layers, and activation function parameters. Evaluation was performed using accuracy, precision, recall, and F1-score metrics. The best results were obtained in the 80:20 data split scenario, with a learning rate of 0.0001, 10 epochs, 2 hidden layers, and a sigmoid activation function, with an accuracy of 97.07%, precision of 97%, recall of 97.14%, and an F1 score of 97.07%. These results indicate that the combination of the HOG and BPNN methods can improve the accuracy of tuberculosis detection and can be the basis for the development of an image-based medical diagnosis support system.

Item Type:

Thesis (Undergraduate)

Contributors:

Contribution	Contributors	NIDN/NIDK	Email
Thesis advisor	Sari, Anggraini Puspita	NIDN0716088605	anggraini.puspita.if@upnjatim.ac.id
Thesis advisor	Al Haromainy, Muhammad Muharrom	NIDN0701069503	muhammad.muharrom.if@upnjatim.ac.id

Subjects:

T Technology > T Technology (General)

Divisions:

Faculty of Computer Science > Departemen of Informatics

Depositing User:

Ananda Ayu Puspitaningrum

Date Deposited:

05 Dec 2025 08:12

Last Modified:

05 Dec 2025 08:12

URI:

https://repository.upnjatim.ac.id/id/eprint/48088