基于圆形均分法耦合双重制约的图像伪造检测算法  

Image Forgery Detection Algorithm Based on Circle Sharing Method and Double Restriction

作　　者：张晓琪[1] 侯世英[2] ZHANG Xiao-qi;HOU Shi-ying(Department of Electronic Information Engineering,Nanchong Vocation and Technology College,Nanchong,Sichuan 637000,China;College of Electrical Engineering,Chongqing University,Chongqing 400044,China)

机构地区：[1]南充职业技术学院电子信息工程系,四川南充637000 [2]重庆大学电气工程学院,重庆400044

出　　处：《西南师范大学学报（自然科学版）》2018年第1期47-53,共7页Journal of Southwest China Normal University(Natural Science Edition)

基　　金：重庆市自然科学基金资助项目(2013BA6017)

摘　　要：提出了基于圆形均分法耦合双重制约的图像伪造检测算法采用Forstner算子对图像中的特征点进行检测,使得算法的检测精度得以提升.利用Haar小波响应值,设计圆形均分法,对SURF生成特征描述子进行改进,改善算法的检测效率.随后,利用特征点及其对应的特征向量,构造双重制约模型,对特征点进行正确匹配.最后,利用余弦度量规则对特征点进行归类,完成对图像的伪造检测.实验测试结果表明,与当前图像伪造检测算法相比较,本文算法具有更高的检测精度以及较强的鲁棒性.A new image forgery detection algorithm based on the circle sharing method and the double restriction has been proposed in this paper.Feature points detection accuracy of the Forstner operator has been applied to detect the image of the feature points in order to improve the detection accuracy of the algorithm.Using the Haar wavelet response value design of circular average methods,methods of generating feature descriptor for SURF have been improved to improve the efficiency of detection algorithm.The feature points and their corresponding feature vectors have been used to construct the dual control model to match the feature points.The feature points have been classified by the cosine metric rule,and the image forgery detection is completed.To test the effectiveness of the algorithm design of the simulation experiment,the simulation results show that with the current image forgery detection algorithm,the proposed algorithm has higher detection accuracy and strong robustness.

关 键 词：图像伪造检测 FORSTNER算子 圆形均分法 双重制约 SURF 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]