Sistem Kontrol Pintu Pintar Berbasis Pengenalan Wajah Serta Gestur Kepala Menggunakan LSTM

Kusuma, Rifan Dana (2025) Sistem Kontrol Pintu Pintar Berbasis Pengenalan Wajah Serta Gestur Kepala Menggunakan LSTM. Other thesis, Institut Teknologi Sepuluh Nopember.

[thumbnail of 5024211011-Undergraduate_Thesis.pdf] Text
5024211011-Undergraduate_Thesis.pdf - Accepted Version
Restricted to Repository staff only

Download (34MB) | Request a copy

Abstract

Sistem keamanan cerdas modern memerlukan mekanisme autentikasi adaptif untuk mengatasi ancaman spoofing yang terus berkembang. Penelitian ini memperkenalkan sistem akses pintu cerdas yang mensinergikan pengenalan wajah dan analisis gerakan kepala berbasis LSTM untuk membangun kerangka otentikasi dua faktor yang kuat. Sistem ini memanfaatkan alur sekuensial: pertama, fitur wajah diekstraksi melalui penyematan gradien berorientasi histogram untuk verifikasi identitas, diikuti oleh jaringan LSTM yang menguraikan pola pergerakan kepala temporal (mengangguk, dan mendongak) sebagai biometrik perilaku dinamis. Dengan mengintegrasikan atribut wajah statis dengan gerakan berbasis gerakan, sistem ini mengurangi kerentanan yang melekat pada sistem biometrik modal tunggal, khususnya terhadap serangan spoofing statis. Arsitektur LSTM, yang dilatih berdasarkan dataset custom yang dibuat secara mandiri oleh penulis menunjukkan performa yang cukup baik hingga mendekati akurasi 100%. Validasi eksperimental terhadap beberapa pengguna menunjukkan tingkat keberhasilan autentikasi keseluruhan yang mirip, mampu bersaing sistem pengenalan wajah konvensional degan persentase keberhasilan sebesar 91% dalam kemampuan anti-spoofing. Diimplementasikan pada platform smarthome hub yang dibuat secara mandiri juga, solusi ini beroperasi dengan biaya relatif lebih rendah dibandingkan alternatif komersial dengan tetap mempertahankan respons waktu nyata (kurang dari 50 ms). Karya ini memadukan sifat biometrik spasial dan temporal melalui LSTM, menawarkan cetak biru terukur untuk sistem pintu pintar yang aman namun dapat diakses di lingkungan perumahan dan IoT.
========================================================================================================================
Modern intelligent security systems require adaptive authentication mechanisms to counter the evolving threat of spoofing. This study introduces a smart door access system that synergizes facial recognition and LSTM-based head movement analysis to establish a robust two-factor authentication framework. The system utilizes a sequential workflow: first, facial features are extracted through histogram of oriented gradients embedding for identity verification, followed by an LSTM network that deciphers temporal head movement patterns (nodding and looking up) as dynamic behavioral biometrics. By integrating static facial attributes with motion-based gestures, the system reduces vulnerabilities inherent in single-modal biometric systems, particularly against static spoofing attacks. The LSTM architecture, trained on a custom dataset independently created by the author, demonstrates strong performance with accuracy approaching 100%. Experimental validation across multiple users shows a comparable overall authentication success rate, rivaling conventional facial recognition systems with an anti-spoofing success rate of 91%. Implemented on a self-built smart home hub platform, the solution operates at a relatively lower cost than commercial alternatives while maintaining real-time response (less than 50 ms). This work combines spatial and temporal biometric characteristics through LSTM, offering a scalable blueprint for secure yet accessible smart door systems in residential and IoT environments.

Item Type: Thesis (Other)
Uncontrolled Keywords: Autentifikasi, Sistem Kontrol Pintu Pintar, Pengenalan Wajah, LSTM, IoT, Authentification, Smart Door Control System, Face Recognition, LSTM, IoT
Subjects: T Technology > T Technology (General) > T57.5 Data Processing
T Technology > T Technology (General) > T58.5 Information technology. IT--Auditing
T Technology > T Technology (General) > T59.7 Human-machine systems.
T Technology > TA Engineering (General). Civil engineering (General) > TA1573 Detectors. Sensors
T Technology > TA Engineering (General). Civil engineering (General) > TA158.7 Computer network resources
T Technology > TA Engineering (General). Civil engineering (General) > TA1637 Image processing--Digital techniques. Image analysis--Data processing.
T Technology > TA Engineering (General). Civil engineering (General) > TA1650 Face recognition. Optical pattern recognition.
T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK5103.2 Wireless communication systems. Two way wireless communication
T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK5105.546 Computer algorithms
Divisions: Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Computer Engineering > 90243-(S1) Undergraduate Thesis
Depositing User: Rifan Dana Kusuma
Date Deposited: 24 Jun 2025 07:10
Last Modified: 24 Jun 2025 07:10
URI: http://repository.its.ac.id/id/eprint/119247

Actions (login required)

View Item View Item