VQ-based algorithms extended to non-embedded watermarking for multimedia ownership prevention systems

This paper presents some novel watermarking approaches that are based on vector quantization and provide ownership protection in gray-level images. Nowadays, information sharing in P2P architecture is immensely involved with knowledge-learning on the Internet. Copyright protection is therefore considered in the studies of P2P systems. The proposed approaches combine images, watermarks, and codebooks together by generating key streams to claim ownership and guarantee the integrity of images. Non-embedded approaches that generate claimed key streams have the benefit of maintaining the authorized image in its original state without any modifications. Compared to Lin et al.'s method, our approaches save both running time and space for storage. Moreover, in order to provide broader applications of image protection, robust watermark and fragile watermark strategies are proposed. The experimental results show that these strategies are effective.