Image splicing detection using Gaussian or defocus blur

Tampering images to create false ones using photo manipulation software is very easy. Tampering may be in the form of image splicing. Image Splicing involves the introduction of a selected region from one image into another with the aim to alter its content. Splicing may be in the form of blurred regions introduced via Gaussian blur. In this paper, a method based on inconsistencies in Gaussian blur is used to test the authenticity of an image. Gaussian blur of an image is first evaluated and standard deviation obtained is used to de blur the image. Forged region with different value of standard deviation are bound to show ringing effects, which highlight the forgery.