A low-cost method for reliable ownership identification of medical images using SVM and Lagrange duality
作者：
Highlights：
• We proposed a method for reliable and low-cost ownership identification in medical image watermarking using the SVM and Lagrange duality.
• The proposed method identified the rightful owner using an approximation approach versus a direct correlation view.
• We examined standard medical images, and successfully identified the rightful owner.
• Context similarities and watermarking attacks result in false positive and false negative errors.
摘要
•We proposed a method for reliable and low-cost ownership identification in medical image watermarking using the SVM and Lagrange duality.•The proposed method identified the rightful owner using an approximation approach versus a direct correlation view.•We examined standard medical images, and successfully identified the rightful owner.•Context similarities and watermarking attacks result in false positive and false negative errors.
论文关键词：Medical image watermarking,Support vector machine,Ownership identification,False positive error,False negative error,Spatial domain
论文评审过程：Available online 25 July 2013.
论文官网地址：https://doi.org/10.1016/j.eswa.2013.07.062
