Authors: Ashish Kumar Chakraverti; Vijay Dhir
Addresses: IKG-Punjab Technical University, Jalandhar, Punjab, India; ABES Institute of Technology, Ghaziabad, UP, India ' M.K. Group of Institution, Amritsar, Punjab, India
Abstract: In this paper, we have proposed a new hybrid approach to detect the cloned object in images. In copy-move forgery, images are forged by making the clone of objects of the same image within the image. In our proposed work, we have integrated two states of art techniques named adhoc method and principle component analysis (PCA) based scale invariant feature transform (SIFT) method. In two step process, we have preprocessed the object image by the proposed dynamic block size technique based local contrast modification and contrast limited adaptive histogram equalisation (DBST-LCM CLAHE) up to desired PSNR and grey level and then after integrated adhoc, and PCA based SIFT algorithm to design a hybrid approach. Proposed method tested over CoMoFoD image database. Proposed method shows the better performance in comparison of the state of art techniques regarding false positive rate (FPR), true positive rate(TPR), and PSNR.
Keywords: image processing; image enhancement; ad hoc; peak signal to noise ratio; PSNR; image forgery; scale invariant feature transform; SIFT; principle component analysis; PCA; contrast limited adaptive histogram equalisation; CLAHE; noise.
International Journal of Forensic Software Engineering, 2019 Vol.1 No.1, pp.3 - 20
Received: 18 Aug 2018
Accepted: 22 Sep 2018
Published online: 28 Jan 2020 *