Image Tamper Detection Algorithm Based On Three-dimensional Reconstruction Technique

Living in the 21 th Century,we have entered an information age of rapid development of science and technology.Smart phones have become a necessary tool in life,through which we can easily access a large number of digital images.It can be said that digital images have been integrated into people's daily life.However,because of the rapid development of image editing tools such as Photoshop ?ACDsee,one can easily modify or synthesize the image content without leaving any visible trace.The authenticity and integrity of the image cannot be guaranteed,and the information carried becomes no longer reliable.In this paper,an image tamper detection method based on 3D reconstruction technology is proposed.After recovering the 3D information of the image to be detected,the 3D spatial relation of the image is deduced,and the object that the 3D realistic scene does not match the 2D plane of the image is an object tampered with.This is based on the fact that the inherent characteristics of the camera imaging process,the content of the image,the relative position relationship of the scene object,the viewpoint and so on,will cause unique traces to the image,and there is a perspective mapping relationship between the real three-dimensional scene and the photos captured by the camera.Because 3D information is scarce in a single image,it is always a difficult problem to reconstruct the 3D structure of the target object in the image.In this paper,three main 3D reconstruction methods of single image are studied:(1)Three-dimensional reconstruction based on statistical learning.It is most important for this algorithm to extract the features that can represent the depth information of the image.Then,the depth information of the image is obtained by establishing a good mathematical statistical model.Finally,the 3D reconstruction is accomplished according to the depth map.(2)Three-dimensional reconstruction based on linear perspective cues.There are a large number of rich geometric objects in the real world,which form a lot of parallel lines,parallel planes,vertical lines and vertical planes.These clues formed by geometric constraints in the projection to the image plane,are capable of completing the selfstandardization of the camera to obtain the depth information of the image.(3)Three-dimensional reconstruction based on shape from shading(SFS).This algorithm is based on the shading of the luminance image to extract the three-dimensional shape of the image,and is the inverse process of the imaging process.The gray level of the image is determined by the shape of the object surface under a certain illumination model and projection model.SFS method is solved by modeling the relationship between the gray value and the surface shape of each point in the image under some ideal assumptions and related constraints.