Boštjan Čargo (2009) Blind detection of duplicate regions in digital images. EngD thesis.
Abstract
This work refers to the research area of digital image processing. Its main purpose is to elucidate the field of automatic digital forgery detection and, within its scope, describe a particular algorithm for blind detection of duplicated image regions: the so-called Duplicate Region Detector (DRD). The algorithm is based on principal component analysis, reduction of image blocks representations, and their lexicographical comparison. Our java implementation was tested on a population with positive and negative examples. Test results enabled us to assess efficiency and efficacy, as well as robustness against specific digital image processing operations. The principal conclusion is a thesis that the algorithm can only be made more efficient by abandoning fixed parameter values that are set in advance in favour of values that are defined based on characteristics of each input image.
| Item Type: | Thesis (EngD thesis) |
|---|
| Keywords: | computer vision, image tampering, duplicated image region, Principal component analysis, reduced representation, candidate list, candidate removal, duplication map |
|---|
| Number of Pages: | 75 |
|---|
| Language of Content: | Slovenian |
|---|
| Mentor / Comentors: | | Name and Surname | ID | Function |
|---|
| prof. dr. Aleš Leonardis | 29 | Mentor |
|
|---|
| Link to COBISS: | http://www.cobiss.si/scripts/cobiss?command=search&base=50070&select=(ID=7018836) |
|---|
| Institution: | University of Ljubljana |
|---|
| Department: | Faculty of Computer and Information Science |
|---|
| Item ID: | 819 |
|---|
| Date Deposited: | 23 Mar 2009 11:48 |
|---|
| Last Modified: | 13 Aug 2011 00:35 |
|---|
| URI: | http://eprints.fri.uni-lj.si/id/eprint/819 |
|---|
Actions (login required)
![[img]](http://eprints.fri.uni-lj.si/style/images/fileicons/application_pdf.png)
