An approach to detecting JPEG down-recompression and seam carving forgery under recompression anti-forensics. (May 2017)

• Record Type: Journal Article
• Authors: Liu, Qingzhong
• Is Part Of: Pattern recognition. Volume 65(2017:May)
• Page Start: 35
• Page End: 46
• Publication Date: 2017-05
• Subjects: Seam carving -- Down-recompression -- Down-scaling -- Image forensics -- Large feature mining; Pattern perception -- Periodicals; Perception des structures -- Périodiques; Patroonherkenning; 006.4
• DOI: 10.1016/j.patcog.2016.12.010
• Languages: English
• ISSNs: 0031-3203

Abstract

Abstract: In multimedia forensics, the detection of forgery on joint photographic experts group (JPEG) images is an interesting and challenging work. JPEG double compression, one of common operations that may occur in the tampering manipulation, has been widely studied. While the quality of the second compression is higher than the quality of the first compression, most approaches have obtained effective detection results, however, it still falls short of accurately detecting the down-recompression that the second compression quality is lower than the first compression quality. Seam carving was originally designed for content-aware image resizing. While it is widely used in computer vision and multimedia processing for legitimate applications, it is also being used for forgery manipulation. Although several methods have been proposed to detect seam carving-based forgery, to this date, the detection of the seam carving forgery under recompression in JPEG images has not been well explored. To address the highly challenging detection problems described above, we proposed a large feature mining-based approach. Over one hundred thousand features from the spatial domain and from the DCT transform domain are developed. Ensemble learning is used to deal with the high dimensionality and to avoid overfitting that may occur with some traditional learning classifier for the detection. Our study demonstrates the efficacy of proposed approach to discriminating JPEG down-recompression and exposing the seam-carving forgery from the same quality and low quality JPEG recompression. And hence, it fills a gap in image forensics. Highlights: The work successfully addresses the challenge in detecting JPEG down recompression. It greatly improves the detection of seam carving forgery under recompression anti-forensics. It sheds light on solving highly challenging multimedia forensics problems by large feature mining.