Rotation Invariant Copy Move Forgery Detection using Discrete Cosine Transform

Digital images occupy a crucial part in various realms of our lifestyle nowadays. However, a digital image may be easily tampered with the help of various image editing applications. Hence digital image forensics holds immense importance among all multimedia-based research topics. The simplest image forgery is Copy-Move Forgery, and for this reason, it is the most prominent forgery method. Several post-processing steps like rotation, blurring, and colour reduction make detecting this kind of forgeries difficult. This paper introduces a novel method of identifying and localizing copy move forgery robust to the rotation of the forged areas in an image. Feature matching using block-based Discrete Cosine Transform and the Euclidean distance between the cumulative block Discrete Cosine Transform values are effective parameters that detect replicated regions and rotation of these regions in the image. The experimental results like accuracy, precision, recall and F1 score provide better intuition into the efficiency of this method compared to the other related works in Copy-Move forgery detection involving rotation detection.