Improvement of real time human face recognition

العناوين الأخرى

تحسين تميز الوجوه خلال الوقت الحقيقي

مقدم أطروحة جامعية

Sulayman, Muna

مشرف أطروحة جامعية

al-Ani, Muzhir Shaban

أعضاء اللجنة

al-Hamami, Ala Husayn
Aqil, Misbah Jumah

الجامعة

جامعة عمان العربية

الكلية

كلية العلوم الحاسوبية و المعلوماتية

القسم الأكاديمي

قسم علم الحاسوب

دولة الجامعة

الأردن

الدرجة العلمية

ماجستير

تاريخ الدرجة العلمية

2009

الملخص الإنجليزي

Face recognition is becoming an increasingly important area with advances in technology and computer vision.

There are many applications that can benefit from face recognition including surveillance systems, human tracking, and police screening.

Researchers have proposed many algorithms for face recognition; these algorithms differ in recognition rate.

This thesis, presents a study of the recognition rate of face recognition algorithms.

This work concentrates on the following algorithms: Principal Component Analysis, Two Dimensional Principal Component Analysis in Column Direction, Two Dimensional Principal Component Analysis in Row Direction and Two Dimensional Two Directional Principal Component Analysis.

All these algorithms are implemented into two stages: training stage and recognition stage.

We present a comparison between the four algorithms with respect to recognition rate.

An important optimization in the training and recognition stages is presented in this work.

In case of optimization, a proposal algorithm is implemented based on Discrete Wavelet Transform, the proposed algorithm minimize the images size.

Then these minimized images are entered to a Principal Component Analysis algorithm.

A complexity reduction is achieved by optimizing the number of comparison operations needed.

This optimization does not increase the recognition rate only, but it is also reduces the execution time, by decreasing the number of comparison operations needed.

This work shows the effect of the optimizations on the recognition rate via applying face recognition algorithms.

These algorithms are implemented using MATLAB programming language.

The testing occurs then using Olivetti Research Laboratory database.

The obtained results show that an improvement of 4 to 5 % is achieved by introducing Discrete Wavelet Transform.

التخصصات الرئيسية

الهندسة الكهربائية

عدد الصفحات

60

قائمة المحتويات

• APA
• MLA
• AMA

لغة النص

الإنجليزية

نوع البيانات

رسائل جامعية

رقم السجل

BIM-526516