基于扩展马尔科夫特征的Seam-Carving篡改检测

针对能够用于图像篡改的Seam-Carving技术，提出了一种基于扩展的马尔科夫特征的Seam-Carving篡改识别算法。该算法充分考虑了Seam-Carving操作导致的图像频域特征的变化，将传统的利用马尔科夫转移概率矩阵求取的图像特征和基于扩展的马尔科夫转移概率特征进行融合，而后利用支持向量机进行分类训练，从而达到有效识别基于Seam-Carving的图像篡改。实验结果表明，提出的方案性能优于传统的基于马尔科夫转移矩阵的特征选择方法以及现有的一些该类图像篡改检测方法。

Seam-Carving forgery detection based on expanded Markov features

To deal with the digital image forgery using Seam-Carving, a detection algorithm based on expanded Markov features was proposed. The algorithm takes full acount of the image frequency domain change caused by Seam-Carving operation, merges the features based on traditional and expands Markov transfer-probability matrix. Trained by SVM, the merged Markov feature can identify the Seam-Carving forgery more effectively. The experiment result shows that the performance of the proposed method is better than that of the method based on traditional Markov features and other existing methods.