Digital image forgery detection

Digital image forgery has become a widespread phenomenon in today’s society. Digital images are easy to manipulate and edit due to availability of powerful image processing and editing software. It is possible to add or remove important features from an image without leaving any obvious traces of ta...

Full description

Saved in:
Bibliographic Details
Main Author: Tay, Qinyuan.
Other Authors: Sudha Natarajan
Format: Final Year Project
Language:English
Published: 2011
Subjects:
Online Access:http://hdl.handle.net/10356/43902
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
Description
Summary:Digital image forgery has become a widespread phenomenon in today’s society. Digital images are easy to manipulate and edit due to availability of powerful image processing and editing software. It is possible to add or remove important features from an image without leaving any obvious traces of tampering. As digital cameras and video cameras replace their analog counterparts, there is a pressing need for authenticating digital images, validating their content and detection of forgeries. The purpose of the project is to conduct an experimental investigation of using SIFT and SURF techniques in digital image detection forgery. The author focuses on the detection of a common digital forgery called the copy-move forgery. A part of the image is replicated and pasted onto a part in the image to remove traces of an important image feature. In this project, the author implements SIFT and SURF techniques to detect the forged part even when the image is altered. The performance of the implementations is described in the later parts of the project.