Watermark embedding using audio fingerprinting

Digital watermarking is a common technique in multimedia security for copyright protection and data authentication. Embedding a digital watermark into a media file is often computationally demanding as multiple operations take place within the process to ensure a high level of perceived quality of the marked copy and a high robustness of the embedded watermark. State of the art watermarking algorithms require time consuming spectral transformation operations as well as windowing and perceptual models for masking the embedded watermark. Our new concept is to set up a collection of pre-computed watermarking signals and mix them with the cover signal for fast and simple embedding. To ensure that the watermark signal is well suited for the embedding position with respect to masking, we suggest using audio fingerprinting technology as matching mechanism. Test results show that our approach in able to watermark content using such proposed lookup collection.