Detection of Local Artificial Changes in Images
N. I. Glumov1, A. V. Kuznetsov2
1 Image Processing Systems Institute, Russian Academy of Sciences
2 Image Processing Systems Institute, Russian Academy of Sciences Korolev Samara State Aerospace University
nglu@smr.ru, kuznetsoff.andrey@gmail.com
Keywords: re-sampling, JPEG, duplicates, Fourier-Mellin transform
Pages: 4-12
Abstract
A solution for the problem of detecting local artificial changes (forgery) in images is proposed. Basic properties of images (availability of duplicates, re-sampling, and application of JPEG compression) that can be analyzed to detect changes made in the image are considered. Currently existing methods of detecting the above-mentioned properties satisfy all requirements to photograph authenticity estimation neither in terms of reliability nor in terms of computational complexity. The problem of further development of these methods and improvement of their characteristics for the purpose of possible application for detecting local artificial changes in large-scale images is solved.
|
