Forensic Detection of JPEG Compressed Image

For the detection of JPEG compression (JC) forensics, this paper proposes the feature vector that extracted using the Hu invariant moment of the forged image. The defined seven dimensionality feature vector is trained in an SVM (Support Vector Machine) classifier for the JC detection (JCD) of the forged images. The performance of the proposed JCD is measured with several types of images: unaltered, JPEG compressed, JPEG double compressed and rotated, respectively. Subsequently, the AUC (Area Under Curve) results of the JC classification are 0.9 over on the trained SVM classifier. Thus, it confirmed that the grade evaluation of the proposed scheme is 'Excellent (A)'.