An enhanced fragile watermarking scheme to digital image protection and self-recovery

Recently, Sarreshtedari and Akhaee proposed a source-channel coding approach to digital image protection and self-recovery. This scheme is based on set partitioning in hierarchical trees (SPIHT) encoding and Reed- Solomon (RS) code. The total watermark bit-budget includes three parts, the first part is composed of source encoder output bits used for image content self-recovery, the second part is composed of channel code parity bits used for correcting the errors of source encoder output bits, and the last part is composed of check bits used for detecting tampered areas of image. They claimed that their scheme greatly improves image recovery quality as a consequence of the designed source and channel codes. However, we consider that their scheme is ineffective under the condition of tampering channel parity bits. In fact, when attacker tampers image content, the channel parity bits will be generally changed. In this way, using the wrong channel parity bits to perform channel decoding, it will get wrong source coded data and the tampered image content will not be correctly recovered. Besides, their scheme has no tampering discrimination ability. If attacker only tampers the watermark bits, their scheme will still generate tampered-recognized image blocks list. Aiming to resolve these important design vulnerabilities, we propose an enhanced fragile watermarking scheme to digital image protection and self-recovery in this paper. Experimental results show the effectiveness of our proposed scheme.