Image forgery detection using Color Coherence Vector

Creation and distribution of digital media has become part of daily life. Digital images are used in many fields but authenticity of an image is needed if it is used as evidence in a courtroom. Thus, development of techniques to ensure image authenticity is an active research topic. Passive methods make use of statistical features of images to judge forgery. These methods divide an image into blocks and use extracted features of them to determine the similarity. In this work, we used Color Coherence Vector (CCV) to determine the similarity among blocks. The vector will designate the coherence of the colors in a region. CCV has not been used to detect image forgery before. Experiments show that the method can detect forged regions even if the image is processed by Gaussian Blurring to hide forgery.