Blind Detection Of Fuzzy Operations In Image Stitching

Along with the sustained rapid economic development and social progress,image modification technology also becomes more and more mature.People begin to have certain understanding of image processing software.Some software such as Photoshop CS5,Mito Xiu Xiu and so on are used in our daily life.Image processing software of Photoshop with its characters of rich-content is easy to operate and convenient to fetch.Photoshop has boomed recently and is becoming the main way of editing images.Splicing image is one of the most common methods of image processing,and in the vernacular it is the process of forging the image through the synthesis of clipping,copying and pasting.The splicing method is divided into splicing synthesis of an image and splicing synthesis of two or more images.But theyare handled in a similar way.The splicing area is handled by moving,deforming,adjusting brightness or blurring the edge of the region to making the image looks more natural.And Intercepted image content is called the splicing area.It can effectively solve the problems of image edge inconsistencies.In this paper,it mainly works on the blind detection of the blurred edges of splicing image.We propose and develop a new method for detecting image splicing,it is one of the most common methods of image tampering.The forger may make the image visually pleasant by some post-processing operations such as blurring the spliced region boundary.Such operations often make detection of tampered images difficult.Furthermore,many digital images contain the out-of-focus blur regions that make detection more difficult.This paper presents a new method of detecting splicing in out-of-focus images based on the edge width and slope of image.Firstly,based on the characteristics that the edge width of blur image area is larger than that of natural image area,the natural image areas are removed if the width of image edge is less than threshold.Secondly,the artificial blur is all elements' statistical average in the designation blur neighborhood,but the out-of-focus blur does not exist the manual designation blur radius and the blur filter shoot to outward from focus point,so an effective scheme is designed to distinguish the artificial blur and out-of-focus blur by calculating the pixel slope of image edge.Finally,based on edge width and edge pixel slope,the proposed method can locate the tampered area in the image.Experimental results show that the proposed method can identify the splicing image effectively.In addition,aiming to the characteristics of image feather edges and the inconsistency of image edge,a new detection algorithm is proposed in this paper,in which the USM sharpening is applied to the image to make the artificial blur edge part of the feathering image intensified,and the feather radius of the edge pixel is calculated to authenticate the image.Firstly,USM enhancement is performed on the composite image by adjusting the scaling factor of the control enhancement effect to widen the difference between feather blur edge and non-blur edge feature in splicing image.Secondly,by calculating the slope of the image edge pixels,according to the correlation coefficient between the pixels and the local standard deviation obtained image feathering radius.Finally,by looking for similar to radius the edge pixels,locate stitching tampering with areas in the image.Experimental results show that the method is able to detect forged in the image area.