Record Details

PELACAKAN DAN PENGENALAN WAJAH MENGGUNAKAN METODE EMBEDDED HIDDEN MARKOV MODELS

Jurnal Informatika

View Archive Info
 
 
Field Value
 
Title PELACAKAN DAN PENGENALAN WAJAH MENGGUNAKAN METODE EMBEDDED HIDDEN MARKOV MODELS
 
Creator Margono, Arie Wirawan; Faculty of Industrial Technology, Petra Christian University
Gunawan, Ibnu; Faculty of Industrial Technology, Petra Christian University
Lim, Resmana; Faculty of Industrial Technology, Petra Christian University
 
Subject Computer Vision, Object Tracking, CamShift, Face Recognition, Hidden Markov Model.
 
Description Tracking and recognizing human face becomes one of the important research subjects nowadays, where it is applicable in security system like room access, surveillance, as well as searching for person identity in police database. Because of applying in security case, it is necessary to have robust system for certain conditions such as: background influence, non-frontal face pose of male or female in different age and race.
The aim of this research is to develop software which combines human face tracking using CamShift algorithm and face recognition system using Embedded Hidden Markov Models. The software uses video camera (webcam) for real-time input, video AVI for dynamic input, and image file for static input. The software uses Object Oriented Programming (OOP) coding style with C++ programming language, Microsoft Visual C++ 6.0® compiler, and assisted by some libraries of Intel Image Processing Library (IPL) and Intel Open Source Computer Vision (OpenCV).
System testing shows that object tracking based on skin complexion using CamShift algorithm comes out well, for tracking of single or even two face objects at once. Human face recognition system using Embedded Hidden Markov Models method has reach accuracy percentage of 82.76%, using 341 human faces in database that consists of 31 individuals with 11 poses and 29 human face testers.


Abstract in Bahasa Indonesia :

Pelacakan dan pengenalan wajah manusia merupakan salah satu bidang yang cukup berkembang dewasa ini, dimana aplikasi dapat diterapkan dalam bidang keamanan (security system) seperti ijin akses masuk ruangan, pengawasan lokasi (surveillance), maupun pencarian identitas individu pada database kepolisian. Karena diterapkan dalam kasus keamanan, dibutuhkan sistem yang handal terhadap beberapa kondisi, seperti: pengaruh latar belakang, pose wajah non-frontal terhadap pria maupun wanita dalam perbedaan usia dan ras.
Tujuan penelitiam ini adalah untuk membuat perangkat lunak yang menggabungkan sistem pelacakan wajah manusia dengan menggunakan algoritma CamShift dan sistem pengenalan wajah dengan menggunakan algoritma Embedded Hidden Markov Models. Sebagai input sistem digunakan video kamera (webcam) untuk input bersifat real-time, video AVI untuk input bersifat dinamis, dan file image untuk input statis. Pemrograman perangkat lunak menggunakan prinsip pemrograman berorientasi objek (OOP) dengan menggunakan bahasa pemrograman C++, kompiler Microsoft Visual C++ 6.0®, dan dibantu dengan library dari Intel Image Processing Library (IPL) dan Intel Open Source Computer Vision (OpenCV).
Hasil pengujian sistem menunjukkan bahwa pelacakan berdasarkan warna kulit manusia dengan menggunakan algoritma CamShift cukup baik, dalam melakukan pelacakan terhadap satu maupun dua objek wajah sekaligus. Sistem pengenalan wajah manusia menggunakan metode Embedded Hidden Markov Models mencapai tingkat akurasi pengenalan sebesar 82.76%, dengan database citra wajah sebanyak 341 citra yang terdiri dari 31 individu dengan 11 pose, dan jumlah citra penguji sebanyak 29 citra wajah.

Kata kunci: Computer Vision, Pelacakan Objek, CamShift, Pengenalan Wajah, Hidden Markov Model.
 
Publisher Institute of Research and Community Outreach - Petra Christian University
 
Date 2004-05-31
 
Type info:eu-repo/semantics/article
info:eu-repo/semantics/publishedVersion
 
Format application/pdf
 
Identifier http://jurnalinformatika.petra.ac.id/index.php/inf/article/view/15440
10.9744/informatika.5.1.pp. 22-31
 
Source Jurnal Informatika; Vol 5, No 1 (2004): MAY 2004; pp. 22-31
 
Language eng
 
Relation http://jurnalinformatika.petra.ac.id/index.php/inf/article/view/15440/15432
 
Rights Authors who publish with this journal agree to the following terms:Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (See The Effect of Open Access).