scale invariant feature transform in matlab
no vote
Due to the rapid development and effectiveness of image processing software, it is easy to manipulate and modify digital images. So it's hard to see the authenticity of the given image. Today, important features can be added or removed from images without obvious traces of tampering. As digital cameras and video cameras replace analog counterparts and need for authenticating digital images, content analysis forgery will only increase. For digital photos to be used as evidence of legal issues, or distributed in the media, it is necessary to check the authenticity of the image. In this paper, sift on the letter describes the image forgery detection method. In particular, we focus on the detection of special type digital forgery -- copy mobile attack, in a copy moveimage forgery method; a part of the image is copied in different positions of the same image in thenpasted. Based on an improved algorithm, scale invariant features transform (SIFT) is used to detect the forgery of this clone. After this technology is applied to reduce the input image of yielda, key point detection and feature descriptor are applied to match the three-dimensional representation of all key points. This method can let us all know whether a mobile attack occurs. In addition, we also use clustering analysis to find the matching points in givesoutput.