28th Signal Processing and Communications Applications Conference (SIU), ELECTR NETWORK, 5 - 07 Ekim 2020
One of the passive image authentication methods, copy move forgery detection has become an important research area. In this study, a robust forgery detection method was proposed considering the attacks that can be applied to the image. In the method, suspicious source and target regions are detected by matching SIFT keypoints extracted after obtaining the texture information of the image. The feature vectors of the image sub-blocks are obtained by using DCT-phase to detect forged pixels within these regions. The fact that the corresponding blocks have similar feature vectors indicates that the pixels in the block are forged. In the performance analysis, online dataset named GRIP was used and compared with popular studies in the literature.