Tampering Localization in Digital Image Using First Two Digit Probability Features

In this paper, we have used the first digit probability distribution to identify inconsistency present in the tampered JPEG image. Our empirical analysis shows that, first two digits probabilities get significantly affected by tampering operations. Thus, prima facie tampering can be efficiently localized using this smaller feature set, effectively reducing localization time. We trained SVM classifier using the first two digit probabilities of single and double compressed images, which can be used to locate tampering present in the double compressed image. Comparison of the proposed algorithm with other state of the art techniques shows very promising results.