Use of Defocus Blur Consistency to Detect Image Forgery

Abstract

A method of defocus blur consistency to detect image forgery has been introduced in this paper. At each edge pixel is estimated for the local blur this reveals the defocus blur inconsistency. The results that obtained are effective hence have good applications in real world which consists of tampered images. There is one model i.e., defocus model it has a phenomena that blur sizes of the kernel sizes would be similar for the patches that have similar distance and this gives raise to image forgery crimes that takes advantage of dissimilar imaging conditions and possibility of blurring. In case of quality of real and virtual image obtained because of blurring motion and defocusing easily shows the seam between both virtual and real word. So the solution of the problem is video see-through augmented reality, this consists of simulations of the optical system of camera when rendering of the virtual objects takes place.