Image Forgery Detection with Modified Adaptive Over Segmentation Technique and Noise Attacks

With the advancement of PC innovation and picture preparing programming, computerized picture imitation has been progressively simple to perform. In any case, advanced pictures are a well-known wellspring of data, and the dependability of computerized pictures is in this manner turning into a significant issue. Image forgery is current research trends as forgery may lead to bad consequences like misuse of someone's information and cheating frauds. It is important to prevent any kind of forgery of bank documents, pictures or any certificates. In this paper, study and implementation of a new image forgery detection technique is done. The technique proposed is inspired from the existing feature matching and adaptive over segmentation technique, under which noise attack is performed and also improved the performance parameters such as precision, recall and F1 measure values are much near 1 than the existing one technique. Further, a new application of this technique is tested on real Aadhar card for image forgery detection which will be useful for the detection any forgery in the Aadhar card of a person.