SISTEM DETEKSI DAN PENGENALAN WAJAH SEBAGAI SECURITY SYSTEM BERBASIS HYBRID DEEP LEARNING ALGORITHM

RAMADHAN, ACHMAD SYAUGI and Dwijayanti, Suci (2022) SISTEM DETEKSI DAN PENGENALAN WAJAH SEBAGAI SECURITY SYSTEM BERBASIS HYBRID DEEP LEARNING ALGORITHM. Undergraduate thesis, Sriwijaya University.

	Text RAMA_20201_03041381722105.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (10MB) | Request a copy
	Text RAMA_20201_03041381722105_TURNITIN.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (7MB) | Request a copy
Preview	Text RAMA_20201_03041381722105_0030078404_01_front_ref.pdf - Accepted Version Available under License Creative Commons Public Domain Dedication. Download (3MB) | Preview
	Text RAMA_20201_03041381722105_0030078404_02.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (788kB) | Request a copy
	Text RAMA_20201_03041381722105_0030078404_03.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (388kB) | Request a copy
	Text RAMA_20201_03041381722105_0030078404_04.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (939kB) | Request a copy
	Text RAMA_20201_03041381722105_0030078404_05.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (193kB) | Request a copy
	Text RAMA_20201_03041381722105_0030078404_06_ref.pdf - Bibliography Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (419kB) | Request a copy
	Text RAMA_20201_03041381722105_0030078404_07_lamp.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (6MB) | Request a copy

Abstract

Currently, the development of security system technology has used a combination of biometrics and artificial intelligence through face detection and recognition systems. Although many studies related to face detection have been carried out, face detection technology still faces many challenges in practical application, where face detection is currently still carried out at the close range and relies heavily on facial features so that the accuracy obtained is highly dependent on the accuracy of feature selection. Therefore, this study developed a real-time face detection and recognition system using a hybrid deep learning algorithm. This system used the performance of the convolutional neural network (CNN) with the FaceNet method and the deep belief network (DBN) algorithm. The objects used in this study were 9 faces, where 6 people become data practitioners who can access the room and 3 people become data unknowns who do not have the right to access the room. The total number of data obtained is 18000 data. Furthermore, the data was trained using FaceNet and DBN, where the best training model was a model that used a learning rate of 0.1 with a loss value of 0.971691. Furthermore, the model was tested on offline and real-time test data. Faces recorded by CCTV were detected by the system using the FaceNet model that has carried out the embedding process and the system were able to recognize faces using the trained DBN model. The test results showed the system can detect and recognize faces in still images offline with a probability of 100%. Meanwhile, in real-time test results, the system detected faces and send the information to users directly via telegram media.

Item Type:	Thesis (Undergraduate)
Uncontrolled Keywords:	PENDETEKSI DAN PNGENALAN WAJAH
Subjects:	T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7885-7895 Computer engineering. Computer hardware
Divisions:	03-Faculty of Engineering > 20201-Electrical Engineering (S1)
Depositing User:	Achmad Syaugi Ramadhan
Date Deposited:	02 Aug 2022 03:52
Last Modified:	02 Aug 2022 03:52
URI:	http://repository.unsri.ac.id/id/eprint/75117

Actions (login required)

View Item