A copy-move forgery detection approach based on local intensity order pattern and patchmatch

In recent years, as the technology improves, every discipline like public, medical, military and forensic has digital images. digital images that are referenced in almost every area can be easily falsified by the use of image editing software by malicious people. Most common forgery method on images is copy-move forgery. In this work, a block based method is proposed in order to detect copy-move forgery. Fetaures extracted from blocks by using Local Density Row Pattern (LIOP) which is a new and more efficient method is added to PatchMatch algorithm in order to detect copy-move forgeries in a faster way. Moreover these are compared with recent works and Attack resistance is tested. Experimental results have reported performance that proposed algorithm detects copy-move forgery even if attacks are noise, blurring, rotation, scaling and JPEG compressing.