Robust Prediction of Copy-Move Forgeries using Dual-Tree Complex Wavelet Transform and Principal Component Analysis

Copy-Move Forgery (CMF) reorders a square in an image to conceal or add information. To distinguish between changing regions, technologies that reveal falsification are needed. This work uses dual tree complex wavelet transform (DTCWT) to deteriorate pictures for copy-move forgery detection (CMFD). After degradation, high sub-band images are partitioned into overlapping squares, and PCA is used to compute features for each block. Then, Euclidean separation distance is utilised to calculate the square block distance between consecutive pairs of square blocks to match similar pairings and indicate possible change zones. The proposed work is evaluated using UCID. The quantitative findings were compared by comparing recall, accuracy, True Positive Rate (TPR), and False Positive Rate (FPR). The suggested strategy’s photos are subjectively inspected. The findings show that the suggested strategy outperforms current strategies available in literature on all criteria.