A Scheme of Digital Image Copy-move Forgery Detection Based on Pseudo-Zernike Moments
The rapid development of image editing technology has made it difficult to detect traces of forgery in professionally processed tampered images,posing a serious threat to the fields of news disscmination,judicial forensics,and information security.Region tampering de-tection schemes for images are mainly used to detect the presence of suspected similar image blocks in an image.Currently,traditional detection strategies are not robust to tampered regions where geometric attacks occur.For the geometric attack scenario,this paper introduces a new de-tection method for copy-paste forged images,which utilizes pseudo-Zernike moment features for feature extraction.By matching the features in neighboring regions,the possible forged re-gions in the image can be detected.Experimental results show that the method not only has good detection effect on conventional attacks(e.g.,add noise,JPEG compression),but also performs well in dealing with geometric attacks such as rotation.These results demonstrate the innovation and wide applicability of the method.
NETL
NSTL
万方数据
