A Generation of Meaningfully Secured Images for Copyright- and Privacy-Protected Image Trading Systems Using Singular Component Interchange
Wannida Sae-Tang, Hitoshi Kiya · 2017
This paper proposes a generation of meaningfully secured images using singular component interchange for copyright- and privacy-protected image trading systems. In the conventional systems, the visually encrypted images are suspicious, because they are not meaningful. Therefore, it can lead the trusted third parties (TTPs) to attack the images. On contrary, in the proposed method, singular values of a commercial image are interchanged with those of a dummy image so that the TTP recognizes the dummy image instead. Experimental results show that the proposed method is superior to the conventional methods in terms of visual protection, image quality, and fingerprinting performance.