A Robust Algorithm for the Detection of Cloning Forgery

Today's era is of the digital images, what we think, imagined is created in the form of digital images. The forgery detection is technique which can detect the cloning from the images. The availability of different images, editing tools have made it easier for a common user to mislead the images and create digital fake images from it. The forgery detection has various techniques like active and passive techniques, moreover the properties like analysis, classification etc, are used to identify the forgery. The work presented in this paper is based on cloning forgery detection. The cloning forgery technique is one of the type of forgery detection techniques in which some area of the image is copied and moved to some another area of the same image which is not easy for the naked eyes to detect easily. The PCA (Principle Component Analysis) technique is used to find the mismatched pixels from the image. In the presented work, the GLCM (Grey Level Co-occurrence Matrix) technique is applied with the PCA for the forgery detection. The proposed work is implemented in MATLAB and results are analyzed in terms of PSNR, Recall, Precision and accuracy. In the presented work, the results are better due to less fault rate or able to recover two different attacks (Gaussian noise and motion blur). It is analyzed that the proposed algorithm performs well as compared to the existing algorithm. In existing method the results of their parameters like accuracy were less accurate then proposed method. Moreover, the proposed method achieves more accuracy precession and recall.