Photographic Composite Detection Using Circles

With the development of computer network and multimedia technology, digital images are more and more available in our daily life and work. However, the development of image processing tools also makes forged images into our view, which challenges the trust of people in the social public information. Therefore, how to judge the authenticity of a given image simply and accurately has become a key problem to be solved.This paper aims to find a solution based on blind detecting to the above problem, using the circles as the clues. That is detecting image authenticity with only the geometric constraints which the image possesses. In this work, we propose three new methods for detecting photographic composites using circles. In particular, we focus on three kinds of scenes: (i) two or more coplanar circles with the same radius; (ii) a single circle with its discriminable center; (iii) a single circle with geometric constraints for camera calibration. For two circles’situations, we first estimate the focal length based on the equality of the sizes of two coplanar circles, and then estimate the normal vector of the world circle plane(each circle determines a normal vector). Inconsistencies in the angles among the normal vectors are used as evidence of tampering. On the other hand, for the single circle case, we warp the circle to make metric measurement. The length ratio of semi-minor axis and semi-major axis of the recti?ed ellipse is used as evidence of forgery. To demonstrate the effectiveness of the approach, we present results for synthetic and visually plausible composite images.