{"title":"基于边缘特征的支持向量机高斯滤波伪图像检测","authors":"K. Rhee","doi":"10.1109/ICUFN.2017.7993782","DOIUrl":null,"url":null,"abstract":"For a design of the Gaussian filtering (GF) detection (GFD) in the tampered digital images, this paper presents three kinds of the new feature vector which are extracted from the edge ratios and the parameters of Hough peaks. In the proposed algorithm, the formed 10-dim. feature vector is trained in SVM (Support Vector Machine) for the GFD. In the experiment, the performance of the proposed GFD scheme is measured in GFw (window = 3D {3 × 3, 5 × 5, compound (3 × 3, 5 × 5)}, σ = 3D 0.5) images versus the median filtering (MF3: window = 3D 3 × 3), the original (ORI), the average filtering (AVE3: window = 3D 3 × 3), and the JPG90 (Quality Factor = 3D 90) images, respectively. However, the measured performances of the AUC by the sensitivity (TP: True Positive rate) and 1-specificity (FP: False Positive rate) is above 0.9. Thus, it is confirmed that the grade evaluation of the proposed algorithm is rated as “Excellent (A).”","PeriodicalId":284480,"journal":{"name":"2017 Ninth International Conference on Ubiquitous and Future Networks (ICUFN)","volume":"92 2 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2017-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"5","resultStr":"{\"title\":\"Forgery image detection of Gaussian filtering by support vector machine using edge characteristics\",\"authors\":\"K. Rhee\",\"doi\":\"10.1109/ICUFN.2017.7993782\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"For a design of the Gaussian filtering (GF) detection (GFD) in the tampered digital images, this paper presents three kinds of the new feature vector which are extracted from the edge ratios and the parameters of Hough peaks. In the proposed algorithm, the formed 10-dim. feature vector is trained in SVM (Support Vector Machine) for the GFD. In the experiment, the performance of the proposed GFD scheme is measured in GFw (window = 3D {3 × 3, 5 × 5, compound (3 × 3, 5 × 5)}, σ = 3D 0.5) images versus the median filtering (MF3: window = 3D 3 × 3), the original (ORI), the average filtering (AVE3: window = 3D 3 × 3), and the JPG90 (Quality Factor = 3D 90) images, respectively. However, the measured performances of the AUC by the sensitivity (TP: True Positive rate) and 1-specificity (FP: False Positive rate) is above 0.9. Thus, it is confirmed that the grade evaluation of the proposed algorithm is rated as “Excellent (A).”\",\"PeriodicalId\":284480,\"journal\":{\"name\":\"2017 Ninth International Conference on Ubiquitous and Future Networks (ICUFN)\",\"volume\":\"92 2 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2017-07-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"5\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2017 Ninth International Conference on Ubiquitous and Future Networks (ICUFN)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/ICUFN.2017.7993782\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2017 Ninth International Conference on Ubiquitous and Future Networks (ICUFN)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICUFN.2017.7993782","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 5
摘要
针对篡改后数字图像的高斯滤波检测设计,提出了从图像边缘比和霍夫峰参数中提取三种新的特征向量。在提出的算法中,形成了10-dim。在支持向量机(SVM)中训练特征向量。在实验中,对GFw (window = 3D {3 × 3,5 × 5,复合(3 × 3,5 × 5)}, σ = 3D 0.5)图像与中值滤波(MF3: window = 3D 3 × 3)、原始(ORI)、平均滤波(av3: window = 3D 3 × 3)和JPG90 (Quality Factor = 3D 90)图像的性能进行了测试。然而,通过灵敏度(TP:真阳性率)和1特异性(FP:假阳性率)测量的AUC性能均在0.9以上。因此,确定本文算法的等级评价为“优秀(A)”。
Forgery image detection of Gaussian filtering by support vector machine using edge characteristics
For a design of the Gaussian filtering (GF) detection (GFD) in the tampered digital images, this paper presents three kinds of the new feature vector which are extracted from the edge ratios and the parameters of Hough peaks. In the proposed algorithm, the formed 10-dim. feature vector is trained in SVM (Support Vector Machine) for the GFD. In the experiment, the performance of the proposed GFD scheme is measured in GFw (window = 3D {3 × 3, 5 × 5, compound (3 × 3, 5 × 5)}, σ = 3D 0.5) images versus the median filtering (MF3: window = 3D 3 × 3), the original (ORI), the average filtering (AVE3: window = 3D 3 × 3), and the JPG90 (Quality Factor = 3D 90) images, respectively. However, the measured performances of the AUC by the sensitivity (TP: True Positive rate) and 1-specificity (FP: False Positive rate) is above 0.9. Thus, it is confirmed that the grade evaluation of the proposed algorithm is rated as “Excellent (A).”