Neural Network Approach to Bio-watermarking for Image Authentication in DCT Domain
Mya Thidar Kyaw, Kyi Soe · 2013
Security has become one of the most significant and challenging problems for spreading new information technology. Security requirements are met by security measures, which generally consist of several security mechanisms such as cryptographic mechanisms and digital watermarking techniques. An image authentication by combination of Online signature and watermarking techniques is proposed in this paper. This paper presented image authentication of two watermarking techniques such as partially blind watermarking technique with Neural Network, and efficient blind watermarking technique with Neural Network based on Discrete Cosine Transform (DCT). In our proposed system, Online handwritten signature is captured by digital input device such as G-Pen F350 tablet and is used as watermark. Partially blind watermarking technique with Neural Network was implemented by sequences of network and efficient blind watermarking technique with Neural Network does not focus on frequency coefficient values. Finally, several performance analysis such as robustness of our secret information against various attacks will be presented. The proposed system can be applied for copyright protection for authorized user using handwritten Online signature.