Improved Feature Vector of Median Filtering Residual for Image Forensics
Kang Hyeon Rhee, Ilyong Chung · 2018
For the detection of median filtering (MF) forensics, this paper proposes an improved feature vector that consists of the refined feature vector of the MFR AR (Median Filter Residual Autoregressive Model) and an additional feature vector about the gradient and edge lines of an image. The improved nine dimensionality feature vector is trained in a SVM (Support Vector Machine) classifier for the median filtering detection (MFD) of the forged images. The performance of the proposed MFD scheme is measured with several types of the operated images: median filtered, unaltered and JPEG compressed, respectively. Subsequently, the AUC (Area Under Curve) results of the proposed MFD scheme have 0.99 over on the trained SVM (Support Vector Machine) classifier. Thus, it confirmed that the grade evaluation of the proposed scheme is `Excellent (A)' in terms of the AUC evaluation.