Ahmed AK. TAHIR


35-52 HUMAN BIOMETRICS AND BIOMETRIC RECOGNITION SYSTEMS; AN OVERVIEW

BIOMETRICA UMANĂ ȘI SISTEMELE DE RECUNOȘTERI BIOMETRICE; O IMAGINE DE ANSAMBLU

Biometric technology is gaining a significant role in presenting the solutions for many issues in various applications that demands person identification such as forensic science, security, finance affairs, border checking and government ministries and offices. It is defined as the technology of analyzing physiological and behavioral traits such as face, fingerprint, iris, retina, voice, and signature etc., for person identification and authorization. Nowadays, lots of research works are carrying out to accomplish biometric recognition systems based on various types of human traits. To provide a comprehensive survey, this paper presents an overview to five biometric traits (iris, fingerprints, face, voice and signature). The overview will cover the way of acquisition, application area, methods of implementation, strength/weakness, and system evaluation.

Keywords: Biometrics, Biometric Recognition System, iris, Fingerprints, Finger Vein, Palm Vein, Face Recognition, Gait Recognition, Human Traits, Voice Recognition, Signature Recognition Cuvinte cheie: Biometrie, Sistem de recunoaștere biometrică, iris, amprente, vene de deget, vena palmei, recunoaștere a feței, recunoaștere a trecerii, trăsături umane, recunoaștere vocală, recunoaștere semnătură

Descarcare text


33-27 A NEW METHOD OF EYELID DETECTION FOR IRIS RECOGNITION SYSTEM

O NOUĂ METODĂ DE DETECTARE A PLEOAPELOR
PENTRU SISTEMUL DE RECUNOAȘTERE A IRISULUI
This paper presents a new method for detecting the upper and lower eyelids. The method is called Refine-Connect-Smooth (R-C-S). It consists of three algorithms, Canny edge detector, Refine Edge Map (REM) and Connect and Smooth Edges (CSE). The algorithm is applied after the iris extraction stage therefore any incorrect detection of the eyelid will not decrease the accuracy of iris localization. The application of the method to CASIA database images version-1.0 has shown reasonable accuracy compared to previous methods of eyelid detection. For upper eyelid the accuracy reached 87.3%, while for lower eyelid the accuracy reached 99.2%.
Keywords: iris recognition system, biometric measurement system, eyelid detection, eyelash detection, iris localization, iris boundary detection

Descarcare text


31-51 FUZIUNEA IMAGINII PENTRU ÎMBUNĂTĂȚIREA IMAGINILOR MULTISPECTRALE PRIN SATELIT

IMAGE FUSION FOR RESOLUTION IMPROVEMENT

OF MULTISPECTRAL SATELLITE IMAGES

 

This paper presents a comparative study concerning the techniques of image fusion that aim at the improvement of the spatial resolution of multispectral satellite images for remote sensing applications. Three methods are used, Hue-Saturation-Intensity color transform (HSI), Principal Component Analysis (PCA), and Brovey technique. Two datasets of two satellites having different resolution and different resolution ratio are used, LANDSAT7 and IKONOS. The measurement of Trade-Off-Value, which is a measure of the degree of resolution improvement and spectral property preservation, is used to evaluate the given methods. The results have shown the superiority of the Brovey method for the LANDSAT7 dataset and the superiority of HSI for the IKONOS dataset.

 

Keywords: image fusion, remote sensing, Hue-Saturation-Intensity transform (HSI), Principal Component Analysis (PCA), Brovey technique

Cuvinte cheie: fuziune a imaginii, teledetecție, transformare a culorii nuanță-saturație-intensitate transformata (HSI), Analiza componentei principale (PCA), tehnica Brovey

Descărcare text


30-29 LOCALIZAREA IRISULUI PENTRU SISTEMUL BIOMETRIC DE IDENTIFICARE A PERSOANELOR

LOCATION IRIS BIOMETRIC PERSON IDENTIFICATION
Methods of iris localization are investigated and new methods for pupil
and limbus boundary detection are introduced. Pupil boundary detection is
done by first pre-processed the iris image using exponential stretch and
Laplacian filter. Then the center and the radius are calculated using newly
developed algorithm. This algorithm does not require any threshold value to
determine the initial center since it is specified automatically. For limbus
boundary detection, the integro – differential operator is used. CASIA database
v1.0 is used to test the efficiency of the given algorithms. The accuracy of
localizing the iris correctly reaches 92 %. All the algorithms were implemented
using programming language Turbo C++.
Sunt investigate și introduse noi metode de localizare a irisului și de
detectare a marginii pupilei și a limitei limbusului (marginea corneei).
Detectarea marginii pupilei se face prin procesarea imaginii irisului, folosind
extinderea exponențială a filtrului Laplacian. Apoi centrul și raza sunt calculate
folosind un nou algoritm. Acest nou algoritm nu necesită nici o valoare limită,
pentru a determina centrul inițial al ochiului, din moment ce acesta este
specificat automat. Pentru detectarea limitei limbusului, se folosește operatorul
diferențial – linie integro, (baza de date CASIA), folosit pentru a testa eficiența
algoritmilor dați. Acuratețea localizării irisului ajunge până la 92 %. Toți
algoritmii au fost implementați folosind limbajul de programare TURBO C ++.
Keywords: iris recognition biometric, human traits, security system,
improved appearance, pupil detection, detection of limbus
Cuvinte cheie: recunoașterea biometrică a irisului, trăsături umane,
sistem de securitate, imagine ameliorată, detectarea pupilei, detectarea
limbusului

Descărcare text