【作者】 王波； 孔祥维； 尤新刚；

【Author】 WANG Bo~1 KONG Xiangwei~1 YOU Xingang~(1,2) 1 Information Security Research Center of Dalian University of Technology,Dalian 116024 2 Beijing Institute of Electronic Technology and Application,Beijing 100091)

【机构】 大连理工大学信息安全研究中心； 北京电子技术应用研究所；

【摘要】 数字相机的普及和图像处理软件的应用使得数字图像正面临着被随意篡改和伪造的威胁。针对模糊润饰后的数字伪造图像,本文提出了一种利用色彩一致性进行定位检测的数字图像取证方法。通过从图像块中提取全局色调率,色调变化率以及异常色调率三个特征,来量化描述模糊润饰操作对成像系统引入的图像局部色彩一致性的破坏。利用支持向量机分类器进行原始像素点和模糊像素点的分类,并以此对模糊润饰的伪造图像区域进行定位。实验表明,该方法能够对模糊润饰的图像进行有效检测和定位。更多还原

【Abstract】 The widely used digital cameras and photo editing software bring us a new problem of tampered image.Focusing on the blurring operation in digital forgeries’ creation,a new digital image forensic method using color consistency is proposed for localization the tampering region in the digital forgeries.By extracting three statistical features including the ratio of global hues,hue bias and abnormal hues,the color consistency introduced by imaging pipeline and destroyed by blurring operation is described.The features are then fed to the support vector machine to classify the authentic pixels and forgeries,as well as localizing the tampering regions.Experiments show the accuracy of localizing the blurring regions in the digital image forgeries.更多还原