Protecting biometric data for personal identification

Abstract

This paper presents a new chaotic watermarking and steganography method to protect biometric data, which is sent over the networks for personal identification. Unlike other methods, we utilized two keys, one for encrypting the biometric template before embedding, which makes our method more secure and another for watermark embedding that makes our method more robust. The proposed method does not require original biometric image for the extraction of watermarked data, and can provide high accuracy of extracted watermarked data even under different noises and distortions. Experimental results show that the performance of the proposed method is encouraging comparable with other methods found in the current literature and can be used in a practical system. © Springer-Verlag Berlin Heidelberg 2004.