Articles

MULTIPLE DISCRIMINANT ANALYSIS WITH FUKUNAGA KOONTZ TRANSFOR AND SUPPORT VECTOR MACHINE FOR IMAGE-BASED FACE DETECTION AND RECOGNITION

Main Article Content

Sri Andriati Asri Widyadi Setiawan

Abstract

MULTIPLE DISCRIMINANT ANALYSIS WITH FUKUNAGA KOONTZ TRANSFOR AND SUPPORT VECTOR MACHINE FOR IMAGE-BASED FACE DETECTION AND RECOGNITION a Sri Andriati Asri, bWidyadi Setiawan aElectrical Engineering Dept., Bali State Polytechnic, Bukit Jimbaran, Kuta Selatan, Badung, Bali b Electrical Engineering Dept., Faculty of Engineering Udayana University, Bukit Jimbaran, Kuta Selatan, Badung, Bali 80361 E-Mail: andriati_s@yahoo.com Abstrak Pengenalan wajah dapat diterapkan pada banyak aplikasi potensial, seperti otentikasi identitas, information security, surveillance dan interaksi manusia komputer. Penelitian ini bertujuan membangun perangkat lunak berbasis Matlab untuk deteksi dan pengenalan wajah dengan masukan berupa citra. Sistem yang akan dibangun meliputi deteksi dan pengenalan wajah. Subsistem Deteksi Wajah memakai Principle Component Analysis (PCA) sebagai ekstraksi fitur dan Jaringan Syaraf Tiruan Perambatan Balik sebagai pengklasifikasinya. Pada Subsistem Pengenalan Wajah memakai metode Support Vector Machine salah satu algoritma kecerdasan buatan yang mampu mengklasifikasikan banyak wajah dengan baik. Metode Multiple Discriminant Analysis with Fukunaga Koontz Transform (MDA/FKT) dipakai sebagai ekstraksi fitur. Pelatihan dan pengujian sistem memakai basis data penelitian, dan basis data standar yaitu basis data ORL sebagai pembanding. Rancang bangun Aplikasi Deteksi dan Pengenalan Wajah telah berhasil diselesaikan pada penelitian ini. Subsistem Deteksi Wajah menghasilkan tingkat keakuratan pendeteksian wajah sebesar 99 %. Pada Subsistem Pengenalan Wajah, tingkat pengenalan basis data penelitian (UNUD) 82,76 %, sedangkan tingkat pengenalan pada basis data ORL 97,5%. Kata kunci: Deteksi Wajah, Pengenalan Wajah, Support Vector Machine, Multiple Discriminant Analysis with Fukunaga Koontz Transform. Abstract Face recognition can be applied to many potential applications, such as identity authentication, information security, surveillance and human computer interaction. This research aims to build a Matlab-based software for face detection and recognition application using an image input form. The system consist of face detection and recognition subsystem. Face detection subsystem using PCA as feature extraction and Back Propagation Neural Network as its classifier. In face recognition subsystem using Support Vector Machine as known as one of the good methods in the artificial intelligence algorithm that is able to classify many faces well. Multiple Discriminant Analysis Method with Fukunaga Koontz Transforms (MDA / FKT) is used as feature extraction. Training and Testing database systems using research (UNUD) database, and ORL database as a comparison. Face detection and recognition application has been successfully completed in this research, face detection subsystem produces face detection accuracy rate of 97.95 %, and for face recognition subsystem, the recognition rate is 82.76 % on research (UNUD) database, while the recognition rate on ORL database is 97.5 %. Key words: Face Detection, Face Recognition, Support Vector Machine, Multiple Discriminant Analysis with Fukunaga Koontz Transform

Downloads

Download data is not yet available.

Article Details

How to Cite
ASRI, Sri Andriati; SETIAWAN, Widyadi. MULTIPLE DISCRIMINANT ANALYSIS WITH FUKUNAGA KOONTZ TRANSFOR AND SUPPORT VECTOR MACHINE FOR IMAGE-BASED FACE DETECTION AND RECOGNITION. Jurnal Ilmiah Kursor, [S.l.], v. 7, n. 2, july 2013. ISSN 2301-6914. Available at: <https://kursorjournal.org/index.php/kursor/article/view/43>. Date accessed: 22 nov. 2019.
Section
Articles