Scalable secret sharing of compressed multimedia

Traditional secret image sharing methods have an all-or-nothing property, which is not suitable for applications which require gradual reconstruction. In this paper, we propose a scalable secret image sharing (SSIS) method that provides gradual reconstruction with smooth scalability. Furthermore, we extend this method to videos and propose a scalable secret video sharing (SSVS) method. These two methods are designed for compressed multimedia (i.e. JPEG images and H.264 videos). In both methods, the size of the shadow images (shares) is reduced to an optimal value. Experimental results and analyses show that the proposed methods are computationally and semantically secure.