As digital images become an indispensable source of information, the authentication of digital images has become crucial.Various techniques of forgery have come into existence, intrusive, and non-intrusive.Image forgery detection hence is becoming more challenging by the day, due to the unwavering advances in image processing.Therefore, image forensics is at the forefront of security applications aiming at restoring trust and acceptance in digital media by exposing counterfeiting methods.
The proposed work compares between various feature selection Bowl Balance Ring algorithms for the detection of image forgery in tampered images.Several features are extracted from normal and spliced images using spatial grey level dependence method and many more.Support vector machine and Twin SVM has been used for classification.A very difficult Connecting Wire problem in classification techniques is to pick features to distinguish between classes.
Furthermore, The feature optimization problem is addressed using a genetic algorithm (GA) as a search method.At last, classical sequential methods and floating search algorithm are compared against the genetic approach in terms of the best recognition rate achieved and the optimal number of features.