An information-space model for explaining generalized information hiding schemes

With the sharp increase of requirement of security, various information hiding schemes have been being developed for last two decades. The implementation of traditional information hiding schemes is based on the insensitivity of human visual and the redundancy of cover data. The secret message is embedded by slightly modifying the insignificant part of cover data. These modifications are imperceptible so that the secret message or watermark signal is more secure. However, as the range of applications continues to expand, many schemes no longer hide information in the "embedding" manner such as zero-watermarking. Some schemes with state-of-the-art working mechanism have broken through the traditional definition of information hiding. Thus, in this paper, we introduce a definition of generalized information hiding and establish a corresponding model based on information-space. Various information hiding schemes are mapped to different sub-spaces in the information-space. Thus, a reasonable explanation for watermarking schemes based on public-key and non-embedding information hiding schemes is provided to improve the theory of modern information hiding.