Multimodal Biometric Authentication using PSO based Watermarking

Multimodal Biometrics is the usage of multiple biometric indicators by personal identification systems for identifying individuals. It is applied to secure and authenticate the biometric data, enhance accuracy of recognition and reduce bandwidth. This paper presents a robust multimodal biometric image watermarking scheme using Particle Swarm Optimization (PSO). The key idea is to watermark an individual's face image with his fingerprint image and demographic data. PSO is used to select best DCT coefficients in the face image for embedding the watermark. The objective function for PSO is based on the human visual perception capability and ability of the watermarked image to sustain image processing attacks. Experimental results show that the proposed technique embeds private biometric data securely in another biometric content without affecting the latter's visual quality. The embedding technique ensures, that at the receiver's side when the watermark is extracted, the practical usefulness of both the biometrics remains intact leading to a more secure and reliable system of personal recognition.