An Efficient Forensic Approach for Copy-move Forgery Detection via Discrete Wavelet Transform

In this modern era, due to advancements in image processing applications; image forgery can be easily performed and difficult to locate by the bare eye. Therefore, the digital images and its content are becoming doubtful thereby losing credibility and truthfulness. Therefore, proving the content realism of an image is an imperative matter. Commonly used type of digital image forgery is copy-move that is carried out in the same image by pasting specific region to another semantically similar region. In this study, a technique is proposed for detecting image forgery by using discrete wavelet transform (DWT). Compared to Fourier transform (FT), DWT is more suitable to analyze image contents on edges or abrupt changes in color contrast. The DWT decomposes the image into 4 sub-bands. Approximation sub-band is employed to implement the present method. DWT reduces the size of the image which in turn minimizes the execution time of the algorithm. The experiment results of the proposed technique revealed that this approach is effective in detecting and localizing the copy-moved region. Therefore, the presented technique can be used for CMIF detection and benefits can be obtained in different fields such as the judiciary, media, and crime investigation.