Development of Image Forgery Detection Techniques from the Perspective of Machine Learning and Deep Learning

Abstract

Imaging technology has taken remarkable strides in recent years, and digital images can now be snapped from a range of handheld devices, like the mobile, wearables, tabs and camcorder. These images are shared on social networks day in and out. On the other hand, software tools for photo editing have seen a new rise since the inception of social networks and filtering is just a few clicks away on a smartphone. These modern advancements in image editing tools have eroded the trust of digital images, thereby challenging the integrity and authenticity of the image. These technologies enable any amateur user to enhance their images on the fly that makes an impact and concern in the digital forensics community.Two fundamental questions are to be answered in digital image forensic analysis. First and foremost, is to identify whether the given Image is altered or not? Secondly, does the given image is acquired from the specified source camera model or not? To promote value additions to the solutions on the aforementioned questions and to strengthen the image forensic analysis, two methods are explored hereby namely, Anti-Forensic Contrast Enhancement Detector (AFCED) and Source Camera Identifier (SCI).