Embedding on-line handwriting signature features as non-blind watermarking

Combined with the electronic documents system, a new algorithm is proposed in this paper to verify the sender´s identity by embedding indiscernibly on-line handwriting signature features into an image. After the decomposition of host image by discrete wavelet transform, the watermark is embedded adaptively according to human vision system. Due to the watermark based on signature features demanding a higher robustness than traditional watermarking, the algorithm embeds the on-line handwriting signature features repeatedly so that the errors can be reduced in the process of watermark extraction. Experiments show that this algorithm has good invisibility and strong anti-attack capability. It has achieved the ultimate goal of the design.