An efficient USM sharpening detection method for small-size JPEG image. (April 2020)

• Record Type: Journal Article
• Authors: Wang, Dongping; Gao, Tiegang
• Is Part Of: Journal of information security and applications. Volume 51(2020)
• Publication Date: 2020-04
• Subjects: USM sharpening -- Small-size JPEG image forensic -- Difference matrix -- Sharpening detection; Computer security -- Periodicals; Information technology -- Security measures -- Periodicals; 005.805
• DOI: 10.1016/j.jisa.2020.102451
• Languages: English
• ISSNs: 2214-2126

Abstract

Abstract: USM (Unsharp masking) sharpening is a popular digital image processing operation utilized to pursue better image quality. In recent years, the detection of USM sharpening has attracted attention from image forensics point of view. Some effective methods have been proposed to detect the sharpening forgery in large size image. It is, however, still a challenging task to detect the sharpened images with small size. To address this challenge, an efficient USM sharpening detection algorithm is proposed for small-size JPEG image. Firstly, the RGB image was converted into the YCrCb image and got the Y channel component. Secondly, the Y channel image was converted into the frequency domain by block DCT transform. Then, the image feature was gotten by difference matrix computation. Finally, the feature set was fed into an SVM classifier for training and testing. The experimental results show the superiority of the proposed method over the existing method in sharpening detection in small-size images.