Copy-Move Forgery Detection Using Shift-Invariant SWT and Block Division Mean Features
Digital images are used in courtrooms as evidence. We cannot predict nativity of the image without forensic analysis. Tampering with the image is common nowadays with a lot of online and offline tools. To hide an object in an image, regions of the same image are copied and pasted on that object, and this is known as copy-move forgery. In this paper, we have introduced a technique to detect such type of forgery, known as CMFD. In this technique, the image is pre-processed by converting RGB into YCbCr and then Y channel is decomposed into four components of translation-invariant stationary wavelet transform (SWT). Its LL (approximation) component is then divided into 8 × 8 blocks. Further, from each block, we have taken six mean features which are calculated by dividing each block into four squares and two triangular blocks and put them into feature vector with block location. After sorting these feature vectors into lexicographical order, we get the location of forged regions.
KeywordsCopy-move forgery detection Digital image forgery Feature extraction Translation-invariant Stationary wavelet transform
- 2.English Oxford Living Dictionaries. Retrieved Oct 5, 2017 from https://en.oxforddictionaries.com/definition/tamper.
- 4.Weihai, L., & Yu, N., Yuan, Y. (2008). Doctored JPEG image detection. In Proceedings of IEEE International Conference on Multimedia and Expo (pp. 253–256). https://doi.org/10.1109/icme.2008.4607419.
- 10.Michael, Z., Xingming, S. (2011). DWT-PCA(EVD) based copy-move image forgery detection. International Journal of Digital Content Technology and its Applications, 5.Google Scholar
- 11.Tralic, D., Zupancic, I., Grgic, S. (2017). GM CoMoFoD—New database for copy-move forgery detection. In Proceedings of 55th International Symposium ELMAR-2013. Retrieved Oct 11, 2017 form http://www.vcl.fer.hr/comofod/download.html.