基于稀疏编码的图像隐写检测技术研究

摘要

隐写术自20世纪末被提出后发展迅速,给信息安全和国家安全来了极大挑战.目前,针对图像隐写的检测技术已成为信息安全领域的一个研究热点.利用稀疏编码在图像处理中的特点和优势,提出一种基于稀疏编码的图像隐写检测方法.从原始图片和隐写图片中提取正负样本用于训练稀疏编码字典,采用最大池化算法来构建图像的统计特征,并使用支持向量机作为分类器.实验表明,相比于传统的Markov隐写检测方法,该方法具有更好的检测准确率.%Steganography develops rapidly since its proposal at the end of the twentieth century,and brings great challenges to information security and national security.At present,the detection technology for image steganography becomes a research hotspot in the field of information security.Based on the characteristics and advantages of sparse coding in image processing,an image steganalysis method based on sparse coding is proposed.The positive and negative samples are extracted from the original picture and the steganographic picture to train the sparse coding dictionary,while the maximum pooling algorithm is used to construct the statistical features of the image,and the support vector machine as the classifier.Experiments show that as compared with the traditional Markov steganography method,this method has fairly good detection accuracy.