Self-Embedding Watermarking Scheme Based on MDS Codes

Abstract

This paper proposes a self-embedding watermarking scheme based on Maximum Distance Separable (MDS) codes. The watermark is comprised of the reference-bits and the authentication-bits. The reference-bits is generated by encoding the principal content of all the image blocks. The encoding matrix is derived from the generator matrix of selected systematic MDS code. Based on this encoding method, the reference-bits embedded in an image block will be shared by all the image blocks. Therefore our scheme realize a new reference share mechanism and is immune to the tampering coincidence and the reference waste. Moreover, the maximal tampering rate can be analyzed from the error resilience of the MDS code. On the receiver side, the tampered image blocks can be located by the embedded authentication-bits. As long as the tampering rate is not larger than the maximal tampering rate, the principal content of the tampered image blocks can be recovered perfectly. The restoration is deterministic and the quality of recovered content is constant. Our experimental results demonstrate that the proposed method outperforms the recently state-of-the-art works.