基于圆域分割耦合法则的图像篡改检测算法

为降低图像伪造算法的错误检测率和漏检测率，利用互相关函数(CCF)，设计了基于圆域分割耦合最优相关法则的图像复制-粘贴篡改检测算法。引入FAST算子，计算像素点及其邻点的灰度值，准确提取图像特征点，并利用特征点对应的直方图信息求取其主方向；同时，在该方向上建立特征点的邻域圆，对该圆域进行分割，计算每个分割区域的梯度特征，获取相应的特征向量；利用互相关函数对特征点间的相关程度进行计算，构建最优相关法则，完成特征匹配。利用匹配特征点的特征向量，计算特征点间的欧氏距离，对特征点进行集群，定位复制-粘贴篡改内容，实现伪造检测。实验结果表明：相对已有的伪造检测技术，所提算法具备更高的检测准确率，且对旋转、缩放等内容修改表现出更高的鲁棒性。

Image tampering detection algorithm based on circular segmentation coupled rule

In order to overcome the problem of false detection and missing detection when the threshold is not set properly, an image copy-paste tamper detection algorithm based on the optimal correlation rule of circle segmentation coupling is designed by using Cross-Correlation Function(CCF). The FAST operator is introduced to extract the image feature points accurately by calculating the gray value of the pixels and their adjacent points. Using the histogram information corresponding to the feature points, the principal direction of the feature points is obtained, and the neighborhood circle of the feature points is established in this direction. Through the segmentation of the circle, the gradient features of each segmentation area are obtained, and the feature vectors of the feature points are obtained. The correlation degree between feature points is calculated by CCF to construct the optimal correlation rule to complete feature matching. The Euclidean distance between feature points is calculated by matching the feature vectors of feature points. The feature points are clustered, and the copy-paste tampering content is locked to realize forgery detection. The simulation experiments shows that the detection results of the proposed algorithm for copy-paste tampered images are more accurate than those of the current algorithm for copy-paste tampered images, and it has higher robustness for content modifying such as rotation and zooming.