Robust Biometric (Offline Handwritten Signature) Watermarking and Authentication Algorithm

In this paper a block based robust watermarking technique using wavelet transform has been presented which exploits human visual model for adapting the watermark data to local properties of the host image. The watermark (offline handwritten signature) is embedded to the most significant coefficients at the high frequency bands of the discrete wavelet transform of host image. The watermark strength is chosen so that the perceptual measure of image distortion based on spatial masking is below a predetermined value. A family of typical image distortions functions like Jpeg Compression, filtering, scaling, Gaussian noise is selected and the robustness as well as imperceptibility of the algorithm tested through experimental results. The proposed method is extended to synergistically amalgamate the Euclidian classifier with biometric watermarking for the purpose of authentication.