Micro and macro statistical features based on Singular Value Decomposition (SVD) have been proposed for source cell-phone identification. The performance of the proposed features is evaluated with nai¿ve and informed classifiers for the identification of original images as well as images under several manipulations. The results have been compared to the state-of-the-art and it has been observed that SVD based features are comparable to their counterparts in the literature at reduced complexity.
Published in:
Information Forensics and Security, 2009. WIFS 2009. First IEEE International Workshop on
Date of Conference: 6-9 Dec. 2009