Image Re-sampling Forgery Detection with Ensemble Classifier

Abstract

  Digital Imaging is broadly expanding its prominence in human’s daily life and everyone is capable enough to modify pictures with the assistance of editing software or programming for the noxious plan. These types of images are giving growth to crime in the field of forensics, law enforcement, education, research etc. So, now it’s very much important to detect such alterations/forgery among pictures for legal and criminal issues. So, this work deals in to identify the forged images among original pictures for resampling (rotate or stretch) detection technique in image forgery domain. For achieving this goal the author has designed a modified Markov Features Based Detection Technique. This proposed technique works with different scaling factors for upscaling, downscaling and rotational dataset with bilinear and bicubic interpolation methods. The experimental results show that the modified Markov Features Based Detection Technique has out-performed in comparison with the existing techniques with the highest accuracy reach of 99.95% for resampling forgery detection technique among forged Images.