A New Efficient Forgery Detection Method using Scaling, Binning, Noise Measuring Techniques and Artificial Intelligence (Ai)
Mahesh Enumula1, M.Giri2, V.K. Sharma3
1Mahesh Enumula, Research Scholar, Department of Electrical Communication Engineering, Bhagwant University, Ajmer (Rajasthan), India.
2Dr. M. Giri, Professor, Department of Computer Science and Engineering, Siddharth Institute of Engineering and Technology, Puttur (Karnataka), India.
3Dr. V. K. Sharma, Professor, Department of Electrical Communication Engineering, Bhagwant University, Ajmer (Rajasthan), India.
Manuscript received on 18 July 2023 | Revised Manuscript received on 05 August 2023 | Manuscript Accepted on 15 August 2023 | Manuscript published on 30 August 2023 | PP: 17-21 | Volume-12 Issue-9, August 2023 | Retrieval Number: 100.1/ijitee.I97030812923 | DOI: 10.35940/ijitee.I9703.0812923
Open Access | Editorial and Publishing Policies | Cite | Zenodo | 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: In the market, new, updated editing tools and technologies are available to edit images, and with the help of these tools, images can be easily forged. In this research paper, we propose a new forgery detection technique that estimates the noise on various scales of input images. The effect of noise on input images is also taken into account. Additionally, the frequency of images is altered due to noise. Furthermore, the noise signal is correlated with the original input images, and in compressed images, the quantisation level frequency is also changed due to noise. We partition the input image into M × N blocks. The resized blocks are then processed further, taking into consideration the image colours. The noise value of each block is evaluated at both local and global levels. For each colour channel of the input image, regional and global noise levels are estimated and compared using a binning method. Also, a heat map was measured for each block and each colour channel of the input image, and all these values were stored in bins. Finally, calculate the average mean value of noise from all noise values. With these values, decide whether the input image is a forgery or not. The performance of the proposed method is then compared with that of existing processes.
Keywords: Image Processing, Machine Learning, Retouching, Binning, Forgery, Water Marking.
Scope of the Article: Machine Learning