【作者】 胡玲娜； 蒋铃鸽； 何晨；

【Author】 HU Ling-Na JIANG Ling-Ge HE Chen School of Electronic,Information and Electrical Engineering Shanghai Jiaotong University,200240

【机构】 上海交通大学电子信息与电气工程学院；

【摘要】 针对最低有效位(LSB)数字水印存在性检测,以8位灰度图像为试验对象,提出了基于支持向量机(SVM)算法的LSB数字水印盲检测算法。考虑到水印实际嵌入情况及图像的局部特性,首先对图像进行分割,并分析图块最小生成树(MST)结点的度在水印嵌入前后分布特性的变化情况。接着结合常用统计量提取待测图像特征。最后使用SVM算法对待测图像是否嵌有LSB数字水印进行分类。大量仿真试验表明基于SVM算法的水印盲检测算法对不同类型的待测图像均有较高的检测率,低嵌入率时的检测率仍然可以满足实际应用要求。更多还原

【Abstract】 A new method based on support vector machine(SVM) is proposed to solve the problem of the presence of digital watermarking on the least significant bit(LSB).8bits gray-scale images are chosen as the test objects.Considering the actual embedding situation and the local characteristics image segmentation is performed firstly.Then statistical distribution before and after watermarking of the node degree in the minimum spanning tree(MST) of the segmented block are analyzed.Combining it with common used statistics image features are extracted afterward.Finally SVM algorithm is performed to determine whether the test image has LSB watermarking.Great number of simulation results show that the blind detection based on SVM algorithm has high detection probability on different image formats.The detection probability at low embedding rate is still satisfied the real application.更多还原