| Title | Algoritmi za prepoznavanje lica |
| Author | Kristina Nikolić |
| Mentor(s) | Darija Marković (thesis advisor)
Slobodan Jelić (thesis advisor)
|
| Abstract | U diplomskom radu opisana je primjena algoritama za prepoznavanje lica u samom po-
četku razvoja algoriama i danas. Opisani su glavni pristupi koji se koriste prilikom osmišljavanja
algoritama kao što su psihloški i matematički pristup. Navedene su najbitnije sastavnice
oba pristupa te je dan uvid u općeniti način pristupanja prepoznavanju lica. Algoritmi za
prepoznavanje lica su u radu podijeljeni na dvije glavne kategorije - algoritmi koji koriste
dvodimenzionalni i algoritmi koji koriste trodimenzionalni prikaz lica za prepoznavanje. Za
svaku od kategorija navedena su po tri algoritma. Od dvodimenzionalnih algoritama opisane
su Nixonova metoda, metoda Principal Component Analysis i Linear Discriminant
Analysis. Metode algoritama koji koriste trodimenzionalni prikaz koje su opisane u radu
su trodimenzionalno prepoznavanje bez rekonstrukcije lica, 3D Morphable Model i metoda
bazirana na komponentama. Također, opisan je praktični dio rada u kojem je dana implemenatacija
metode Principal Component Analysis i opisana je baza fotografija korištena u
navedenoj implementaciji. U dijelovima isstraživanja ovoga rada korištena je FERET baza
fotografija lica prikupljena u sklopu FERET programa, sponzorirana od strane ureda DOD
Counterdrug Technology Development Program. Osim toga, pojašnjen je način korištenja
implementiranog algoritma |
| Keywords | Face recognition Nixon method Principal Component Analysis method Linear Discriminant Analysis method three-dimensional model for face recognition without face reconstruction 3D Morphable Model method based on face components |
| Parallel title (English) | Face Recognition Algorithms |
| Committee Members | Kristian Sabo (committee chairperson)
Darija Marković (committee member)
Zoran Tomljanović (committee member)
|
| Granter | Josip Juraj Strossmayer University of Osijek
Department of Mathematics |
| Lower level organizational units | Chair of Applied Mathematics
Applied Mathematics Research Group |
| Place | Osijek |
| State | Croatia |
| Scientific field, discipline, subdiscipline | NATURAL SCIENCES
Mathematics
|
| Study programme type | university |
| Study level | graduate |
| Study programme | Mathematics; specializations in: Financial and Statistical Mathematics, Mathematics and Computer Science, Industrial and Applied Mathematics |
| Study specialization | Mathematics and Computer Science |
| Academic title abbreviation | mag.math. |
| Genre | master's thesis |
| Language | Croatian |
| Defense date | 2016-10-21 |
| Parallel abstract (English) | This thesis describes usage of face recognition algorithms since they first appeared until
today. There is given insight into main approaches used in developing algorithms as psychological
and mathematical approach. The most important components of both approaches
are specified and there are explained general methods of face recognition. Face recognition
algorithms are separated into two categories - algorithms using two-dimensional and
three-dimensional face representations. There are described three methods for each category
of algorithms. Among two-dimensional algorithms there are described Nixon method, Principal
Component Analysis method and Linear Discriminant Method. Three-dimensional
algorithms category consists of explanation od method for face recognition without face reconstruction,
3D Morphable Model method and method based on face components. Further
more, there is explanation of practical part of this thesis where implementation of Principal
Component Analysis method is given and there is description of images database used in
implementation. Portions of the research in this paper use the FERET database of facial
images collected under the FERET program, sponsored by the DOD Counterdrug Technology
Development Program Office. Also, there is explained usage of implemented algorithm. |
| Parallel keywords (Croatian) | Prepoznavanje lica Nixonova metoda metoda Principal Component Analysis Linear Discriminant Analysis metoda trodimenzionalni model za prepoznavanje bez rekonstrukcije lica 3D Morphable Model metoda bazirana na komponentama lica |
| Resource type | text |
| Access condition | Open access |
| Terms of use |  |
| URN:NBN | https://urn.nsk.hr/urn:nbn:hr:126:340618 |
