A secret image sharing scheme based on vector quantization mechanism

Abstract

In this paper, we proposed an image sharing method adopting (t, n) threshold scheme. The secret image goes through Vector Quantization (VQ) scheme and permutes the locations of indices to obtain permuted indexed secret image. We select t indices to form a (t-1)-degree polynomial by taking t indices as its coefficients. Then, the values of blocks' indices in cover image are fed into the (t-1)-degree polynomial to get the output values. After embedding the output values back to the pixels in block, we can obtain meaningful shadow images. Any t out of n shadow images can reconstruct the permuted indexed secret image. By inversing permutation and decoding VQ indices, we can finally reconstruct the secret image. Experimental results show that our method can get high image quality for the resultant shadow images and reconstruct secret images as well. © IFIP International Federation for Information Processing 2006.