Prototype Based Image Forgery Detection Based on Clustering and DWT Transform
Ashish Kumar Sharma1, Shiv Kumar Sahu2, Amit Mishra3
1Ashish Kumar Sharma, Department of Information Technology, Technocrats Institute of Technology, Bhopal (M.P.), India.
2Dr. Shiv K Sahu, Assoc. Professor & Head, Department of Information Technology, Technocrats Institute of Technology, Bhopal (M.P.), India.
3Mr. Amit Mishra, Asst. Professor, Department of Information Technology, Technocrats Institute of Technology, Bhopal (M.P.), India.
Manuscript received on May 12, 2016. | Revised Manuscript received on May 20, 2016. | Manuscript published on June 25, 2016. | PP: 1-5 | Volume-4 Issue-5, June 2016. | Retrieval Number: E1098054516
Open Access | Ethics and Policies | Cite
© The Authors. Published By: Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: In this paper proposed a image forgery detection method. The proposed method is a combination of prototype of clustering and transforms function. The prototype clustering technique gives the patch pattern and wavelet transform gives texture feature. For the texture extraction of image used wavelet transform function, these function is most promising texture analysis feature. For the selection of feature generation of pattern used clustering technique. Clustering technique is unsupervised learning technique process by iteration. The proposed method achieves 100% accuracy in just copy-move forgery (without any change in the size or characteristics of the object) forgery without post-processing and 98.43%, 86.58%, and 95.12% accuracies in copy-move forgery with rotation, scaling, and reflection, respectively.
Keywords: Image forgery detection, Digital images, Photography, Haar Transform, Wavelet, SBD.