KLASIFIKASI FRAKTUR TULANG PADA CITRA RONTGEN MENGGUNAKAN METODE CANNY EDGE DETECTION DAN CONVOLUTIONAL NEURAL NETWORK

SYAFIYA, FIOLINORA and Fachrurrozi, Muhammad (2024) KLASIFIKASI FRAKTUR TULANG PADA CITRA RONTGEN MENGGUNAKAN METODE CANNY EDGE DETECTION DAN CONVOLUTIONAL NEURAL NETWORK. Undergraduate thesis, Sriwijaya University.

	Text RAMA_55201_09021282025111.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (2MB) | Request a copy
	Text RAMA_55201_09021282025111_TURNITIN.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (5MB) | Request a copy
	Text RAMA_55201_09021282025111_0222058001_01_front_ref.pdf - Accepted Version Available under License Creative Commons Public Domain Dedication. Download (590kB)
	Text RAMA_55201_09021282025111_0222058001_02.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (392kB) | Request a copy
	Text RAMA_55201_09021282025111_0222058001_03.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (246kB) | Request a copy
	Text RAMA_55201_09021282025111_0222058001_04.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (466kB) | Request a copy
	Text RAMA_55201_09021282025111_0222058001_05.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (691kB) | Request a copy
	Text RAMA_55201_09021282025111_0222058001_06.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (10kB) | Request a copy
	Text RAMA_55201_09021282025111_0222058001_07_ref.pdf - Bibliography Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (141kB) | Request a copy
	Text RAMA_55201_09021282025111_0222058001_08_lamp.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (93kB) | Request a copy

Abstract

Fractures in bones are a common occurrence in the medical field. Accidents are the leading cause of such injuries in Indonesia, particularly among children and males. Currently, fracture identification through X-ray images is performed using conventional methods that rely on visual analysis by doctors, which are often inaccurate. This study proposes the use of Convolutional Neural Network (CNN) methods to classify bone fractures in X-ray images after edge detection using the Canny Edge Detection method. The system was built by training separately on humerus and forearm bones using 10 different combinations of Canny thresholds and epochs. For the humerus bone, the best result achieved an accuracy rate of 0.91 with a combination of Canny threshold 100-150 and 10 epochs. In contrast, for the forearm bone, there were three combinations with the same accuracy rate of 0.82. Canny threshold 100-150 with 10 epochs, and 150-200 and 200-255 with 25 epochs. Based on these results, this study demonstrates that the humerus bone is more suitable for developing an edge detection system compared to the forearm bone. The similar bone structure but different positions of the bones in the X-ray image dataset greatly affect the final classification results after applying Canny edge detection.

Item Type:	Thesis (Undergraduate)
Uncontrolled Keywords:	Computer Vision, Edge Detection, Convolutional Neural Network
Subjects:	T Technology > T Technology (General) > T1-995 Technology (General)
Divisions:	09-Faculty of Computer Science > 55201-Informatics (S1)
Depositing User:	Fiolinora Syafiya
Date Deposited:	30 Aug 2024 03:23
Last Modified:	30 Aug 2024 03:23
URI:	http://repository.unsri.ac.id/id/eprint/155909

Actions (login required)

View Item