Biometric data-based robust watermarking scheme of video streams

In this paper we propose a method that identifies users in latest compressed streaming with their biometric watermark. Since the fingerprint is the most economical and the most widely used biometric feature, the digital fingerprint watermark system can be one of the best biometric security system. In this paper we propose a digital fingerprint watermark technique that can survive under very low bit rate compression. The proposed algorithm consists of preprocessing of the fingerprint image, watermark insertion using discrete wavelet transform, and watermark recovery. The damaged watermark by high compression ratio can be effectively restored with preprocessing before watermark insertion and accumulation of extracted watermarks in the stream. The simulation results show that the extraction accuracy increases over 99% in the highly compressed H.264 streaming.