Copy-Move Image Forgery Detection Using Discrete Wavelet Transform

To expressing visual information Digital images play a significant role. On social media this information in form of digital image is continually increasing. Now a day easily digital images are misrepresented by peoples because availability of software using for image processing. In this paper we represent novel approaches which identify Copy-Move forgery in image using DWT. First we read image and convert this image into grayscale image. Then apply wavelet transform. Extract overlapped high contrast blocks of the coarsest level. Then lexicographically sorting and calculate phase correlation between rows. Extract blocks from reference and match region. Then draw the duplicate region. For evaluation of our system we calculate TPR, FPR, Accuracy and AUC. Our system accuracy \({\simeq}\) 94.58%.