Veiled-person identification based on part of face and eyes

العناوين الأخرى

التحقق من هوية الأشخاص المنقبين اعتمادا على العينين و جزء من الوجه

مقدم أطروحة جامعية

al-Hasanat, Samahir Madd Allah Muslim

مشرف أطروحة جامعية

al-Mahadin, Bassam
al-Hasanat, Ahmad Bashir

أعضاء اللجنة

Hammuri, Awni Mansur
al-Dhunaybat, Ahmad Abd Allah

الجامعة

جامعة مؤتة

الكلية

كلية تكنولوجيا المعلومات

القسم الأكاديمي

قسم الحاسوب

دولة الجامعة

الأردن

الدرجة العلمية

ماجستير

تاريخ الدرجة العلمية

2017

الملخص الإنجليزي

In modern automated life, biometric applications become an imperative necessity in many life scopes.

Human face is considered as the most user friendly and realistic identifier.

Therefore, face recognition has been a hot and significant research field in last few decades.

Human face is rich in lineaments that ease person recognition, but in restricted cases some of those lineaments are covered by a veil for various causes.

Is the computer system able to identify this veiled-one? Is it able to distinguish veiled-male from veiled-female? These whys and wherefores need to be answered in this thesis.

This thesis aims to investigate to what extents a computer system can identify veiled-human and recognize gender using eyes and the surroundings.

For the purpose of this study, we have created a new database called Veiled-Person Identification (VPI) by a mobile phone camera; VPI consists of 1200 images for 100 different adult veiled-persons (50 males and 50 females).

In all our experiments VPI has been utilized for training and testing.

The proposed method in this thesis is a new method and it is a fusion between Geometrical face features (Edge ratio) and Textural features (Probability Density Function (PDF) of the color moments).

The proposed method has achieved a high success rate for veiled-face identification on VPI database.

Accuracy is ranging from ::.88% to ;9.22% depending on the used classifier.

In addition to, the proposed method achieved up to 99.41% success rate for gender classification.

Finally, the proposed method has been compared with some of wellknown approaches that have been applied in face recognition.

It has exceeded all of them in term of accuracy and dimensionality

التخصصات الرئيسية

تكنولوجيا المعلومات وعلم الحاسوب

الموضوعات

• التعرف على الوجه البشري(علم الحاسوب)

عدد الصفحات

58

قائمة المحتويات

• APA
• MLA
• AMA

لغة النص

الإنجليزية

نوع البيانات

رسائل جامعية

رقم السجل

BIM-749392