Knjižnica Filozofskog fakulteta
Sveučilišta u Zagrebu
Faculty of Humanities and Social Sciences Institutional Repository

Softver za prepoznavanje lica

Downloads

Downloads per month over past year

Urek, Lucija. (2014). Softver za prepoznavanje lica. Diploma Thesis. Filozofski fakultet u Zagrebu, Department of Information Science. [mentor Stančić, Hrvoje and Lopina, Vjera].

[img]
Preview
PDF (Croatian)
Download (1MB) | Preview

Abstract

Topic of this paper is the problem of finding efficient software for face recognition and specific adjacent algorithm. The software is based on photographs of people’s faces which are a template that are then compared with database of different faces, searching for a match. The first step at any face recognition software is face detection, and then comes face tracking and face recognition. It will be clarified how different terms and conditions of obtaining a photograph of people’s faces affect identification, because the faces go through many changes and finding the right software has been a challenge for scientists in the last 20 years. The only similar and currently the best such system, that scientists use as starting point, is human ability of recognizing people i.e. human brain and eyes. Basic starting point for algorithm 5 development is recognition of a human face through a photograph or a video-based recognition, and problem of the design of this algorithm, which takes into consideration all the conditions in which the face is photographed and all the changes of ageing and the way of life, is the field of interest in this paper. One of the most popular algorithms will be explained (Viola – Jones), as well as software usage for different purposes and comparison with other similar systems.

Item Type: Diploma Thesis
Uncontrolled Keywords: Face Recognition Software, face detection, face tracking, face recognition, algorithm
Subjects: Information sciences > Social-humanistic informatics
Departments: Department of Information Science
Supervisor: Stančić, Hrvoje and Lopina, Vjera
Date Deposited: 27 Mar 2015 10:08
Last Modified: 27 Mar 2015 10:08
URI: http://darhiv.ffzg.unizg.hr/id/eprint/5202

Actions (login required)

View Item View Item