In today’s digital world, information is transferred in the form of document, image, audio and video. Digital images have become one of the main methods of communication. An image can define a situation or incident in a better way than words. Images can be captured easily from non-expensive devices. Everyone can capture, store, share and alter an image easily.
Digital images are used in forensic investigation, criminal investigation, medical imaging, journalism and social networking sites such as Facebook, Twitter etc. The usage of digital images and various image editing software’s has increased. Therefore it is easy to edit the image which can change all its semantics. This is called digital image forgery. Human vision cannot discover the forgery in the re-created image.
Therefore it becomes very essential and important to authenticate the originality of an image. There are two approaches to find forgery in digital images. The first approach is active approach. In this approach the authentication of image is done by some information which is pre imbedded in the image such as digital watermark and digital signatures. The second approach called passive approach, which does not require any former information about the image. This is also called Blind approach.
Image forensic is a passive approach which assumes that these digital image forgeries do not leave any visual effect on image, they manipulate the statistics of the image. One of the categories of passive approach are Pixel based image forgery. It works mainly on pixels of an image and detects the statistical differences at pixel level. These are further categorized: Image Resampling, Image Splicing and Copy-Move forgery. In Copy-Move forgery one part of the image is copied and pasted to another part in order to hide some part of the image. There are various algorithms to detect this kind of forgery, which work mainly on JPEG image file format.
The main aim is to study various existing algorithms of detecting digital copy-move image forgery for different types of image formats, analyse these algorithms and suggest whether an algorithm for one image format will be suited for another type of digital image format.