In this paper,we propose a novel coverless image steganographic scheme based on a generative model.In our scheme,the secret image is first fed to the generative model database,to generate a meaning-normal and independ...In this paper,we propose a novel coverless image steganographic scheme based on a generative model.In our scheme,the secret image is first fed to the generative model database,to generate a meaning-normal and independent image different from the secret image.The generated image is then transmitted to the receiver and fed to the generative model database to generate another image visually the same as the secret image.Thus,we only need to transmit the meaning-normal image which is not related to the secret image,and we can achieve the same effect as the transmission of the secret image.This is the first time to propose the coverless image information steganographic scheme based on generative model,compared with the traditional image steganography.The transmitted image is not embedded with any information of the secret image in this method,therefore,can effectively resist steganalysis tools.Experimental results show that our scheme has high capacity,security and reliability.展开更多
The security of steganographic system is significant in this research field. The security defined by relative entropy D (P C || PS )is widely cited to measure the security of different steg-anographic systems. However...The security of steganographic system is significant in this research field. The security defined by relative entropy D (P C || PS )is widely cited to measure the security of different steg-anographic systems. However,two examples are presented to show that some limitations exist in Cachin’s definition. Based on the analysis of a basal hypothesis testing problem,a very useful con-clusion can be drawn to define the security of steganographic system. According to the above illation,an amendatory definition is presented based on the probability of the empirical probability distribution. With the help of new definition,the relationship between security and capacity can be interpreted clearly,and truly secure steganographic method could be designed.展开更多
In order to estimate maximum steganographic capacity of discrete cosine transform（DCT） domain in JPEG image, this paper presents a method based on the maximize capacity under undetectable model（MCUU）. We analyze t...In order to estimate maximum steganographic capacity of discrete cosine transform（DCT） domain in JPEG image, this paper presents a method based on the maximize capacity under undetectable model（MCUU）. We analyze the relation between steganographic capacity and affecting factors（image size, steganography operator, loading band, embedding intensity and image complexity）. Then we design a steganography analyzer architecture and a steganographic algorithm which can dynamically increase the steganographic capacity. Compared with other methods of embedding capacity estimation in DCT domain, the proposed methods utilizes general steganalysis methods rather than one specific steganalysis technique and takes five essential factors into account, which improves the commonality and comprehensiveness of capacity estimation, respectively. The experimental results show that steganographic capacity for quantization index modulation（QIM） is almost twice that of spread spectrum（SS） based on MCUU model.展开更多
自适应隐写是图像隐写方向的研究热点,它通过有效地设计隐写失真函数,自适应地将秘密信息隐藏在图像复杂的纹理区域,具有很强的隐蔽性.近年来,基于生成对抗网络的隐写失真函数设计研究在空域灰度图像上已经取得了突破性的进展,但是目前...自适应隐写是图像隐写方向的研究热点,它通过有效地设计隐写失真函数,自适应地将秘密信息隐藏在图像复杂的纹理区域,具有很强的隐蔽性.近年来,基于生成对抗网络的隐写失真函数设计研究在空域灰度图像上已经取得了突破性的进展,但是目前还没有针对空域彩色图像的研究.与灰度图像相比,彩色图像隐写需要考虑保护RGB通道间相关性,同时合理地分配RGB这3个通道的嵌密容量.设计了一个基于生成对抗网络设计空域彩色图像隐写失真函数的框架CIS-GAN(color image steganography based on generative adversarial network),生成器网络采用两个U-Net子网络结构,第1个U-Net子网络生成修改概率矩阵,第2个U-Net子网络进行正负向修改概率调节,有效地降低对彩色图像通道相关性的破坏.针对彩色图像载体,修改灰度图像隐写分析器作为网络的对抗部分.在生成器损失函数中对彩色图像3个通道总的隐写容量进行控制,生成器能够自动学习分配3个通道嵌密容量.实验结果表明,与现有彩色图像隐写失真函数设计方法相比,提出的网络结构能够更好地抵抗彩色图像隐写分析器的检测.展开更多
基金This paper was supported by the National Natural Science Foundation of China(No.U1204606)the Key Programs for Science and Technology Development of Henan Province(No.172102210335)Key Scientific Research Projects in Henan Universities(No.16A520058).
文摘In this paper,we propose a novel coverless image steganographic scheme based on a generative model.In our scheme,the secret image is first fed to the generative model database,to generate a meaning-normal and independent image different from the secret image.The generated image is then transmitted to the receiver and fed to the generative model database to generate another image visually the same as the secret image.Thus,we only need to transmit the meaning-normal image which is not related to the secret image,and we can achieve the same effect as the transmission of the secret image.This is the first time to propose the coverless image information steganographic scheme based on generative model,compared with the traditional image steganography.The transmitted image is not embedded with any information of the secret image in this method,therefore,can effectively resist steganalysis tools.Experimental results show that our scheme has high capacity,security and reliability.
基金Supported by the National Natural Science Foundation of China (No.60572111)
文摘The security of steganographic system is significant in this research field. The security defined by relative entropy D (P C || PS )is widely cited to measure the security of different steg-anographic systems. However,two examples are presented to show that some limitations exist in Cachin’s definition. Based on the analysis of a basal hypothesis testing problem,a very useful con-clusion can be drawn to define the security of steganographic system. According to the above illation,an amendatory definition is presented based on the probability of the empirical probability distribution. With the help of new definition,the relationship between security and capacity can be interpreted clearly,and truly secure steganographic method could be designed.
基金Supported by the National Natural Science Foundation of China(61170271,61170272,61272310,61373131,61573316)Zhejiang Provincial Natural Science Foundation of China(LY15F020032,LQ12 F02016,LQ15E050006)
文摘In order to estimate maximum steganographic capacity of discrete cosine transform（DCT） domain in JPEG image, this paper presents a method based on the maximize capacity under undetectable model（MCUU）. We analyze the relation between steganographic capacity and affecting factors（image size, steganography operator, loading band, embedding intensity and image complexity）. Then we design a steganography analyzer architecture and a steganographic algorithm which can dynamically increase the steganographic capacity. Compared with other methods of embedding capacity estimation in DCT domain, the proposed methods utilizes general steganalysis methods rather than one specific steganalysis technique and takes five essential factors into account, which improves the commonality and comprehensiveness of capacity estimation, respectively. The experimental results show that steganographic capacity for quantization index modulation（QIM） is almost twice that of spread spectrum（SS） based on MCUU model.
文摘自适应隐写是图像隐写方向的研究热点,它通过有效地设计隐写失真函数,自适应地将秘密信息隐藏在图像复杂的纹理区域,具有很强的隐蔽性.近年来,基于生成对抗网络的隐写失真函数设计研究在空域灰度图像上已经取得了突破性的进展,但是目前还没有针对空域彩色图像的研究.与灰度图像相比,彩色图像隐写需要考虑保护RGB通道间相关性,同时合理地分配RGB这3个通道的嵌密容量.设计了一个基于生成对抗网络设计空域彩色图像隐写失真函数的框架CIS-GAN(color image steganography based on generative adversarial network),生成器网络采用两个U-Net子网络结构,第1个U-Net子网络生成修改概率矩阵,第2个U-Net子网络进行正负向修改概率调节,有效地降低对彩色图像通道相关性的破坏.针对彩色图像载体,修改灰度图像隐写分析器作为网络的对抗部分.在生成器损失函数中对彩色图像3个通道总的隐写容量进行控制,生成器能够自动学习分配3个通道嵌密容量.实验结果表明,与现有彩色图像隐写失真函数设计方法相比,提出的网络结构能够更好地抵抗彩色图像隐写分析器的检测.
