针对图像篡改中常见的边缘模糊操作的鉴别取证，提出一种新的取证算法．该算法首先利用模糊数学对篡改过的人工模糊边缘进行图像模糊增强，得到人工模糊操作边缘被增强而其它部分被减弱的图像; 然后对原篡改图像进行分块，利用粗糙集理论获得约减属性和约减规则对子图像块进行分类，排除平坦区域和自然清晰强边缘图像块后得到另外一个图像; 最后将前面得到的两个图像融合得到结果图像．实验结果表明该算法能较好地排除平坦区域和强边缘而保留人工模糊边缘，从而达到被动取证的目的．

Aiming at the edge blurring manipulation in image tampering,a new algorithm for image forensics is presented. In this algorithm,first,the tampered blurring edge is enhanced based on the fuzzy mathematics,while the real clear edge and flat region is weakened,thus obtaining an image. Next,the tampered image is divided into many sub-images,which are then classified to remove flat regions and real clear sharp edges by using the attribute
reduction of rough set theory,thus obtaining another image. Finally,a fusion of the two above-mentioned images is performed to obtain the resulting image. Experimental results show that the proposed algorithm is an effective passive forensics method that reveals the blurred regions but excludes flat regions and real clear sharp edges.