Exploiting Manipulated Region in an Image using Integrated Convolution Neural Network and LRW Segmentation Features
C. Rajalakshmi1, M.Germanus Alex2, R.Balasubramanian3
1C.Rajalakshmi, Lecturer Department of Computer Science, Manonmaniam Sundaranar University, Tirunelveli Kamarajar Government Arts college Surandai.
2Dr.M.Germanux Alex, Professor & Head in the Department of Computer Science, Kamarajar Government Arts College, Surandai.
3Dr.R.Balasubramanian, Professor & Head in the Department of Computer Science & Engineering Manonmaniam Sundaranar University, Tirunelveli.
Manuscript received on 13 August 2019. | Revised Manuscript received on 19 August 2019. | Manuscript published on 30 September 2019. | PP: 5488-5495 | Volume-8 Issue-3 September 2019 | Retrieval Number: C5097098319/2019©BEIESP | DOI: 10.35940/ijrte.C5097.098319
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. 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: To locate the manipulated region in digital images, we suggest to use Convolution Neural Networks and the segmentation based analysis. A unified CNN architecture is designed with set of training procedures for sampled training patches. Tampering map can be generated for the above said Convolution Neural Networks with the help of tampering detectors. In the other hand, a segmentation using lazy random walk based method is second-hand to generate the tampering chance map, finally integrate the maps and generate the final decision map. This can help to locate the manipulated region accurately. Experiments are conducted using the various datasets to prove the efficiency of the suggest method.
KEY WORDS: Tampering, Segmentation, Forgery, Convolution Neural Networks.
Scope of the Article: Neural Information Processing