Multimodal biometric image watermarking using two-stage integrity verification

  • Authors:

  • Won-gyum Kim;HeungKyu Lee

  • Affiliations:

  • Copyright Protection Center, Seoul, South Korea;Department of Visual Information Processing, Korea University, Seoul, South Korea

  • Venue:
  • Signal Processing
  • Year:
  • 2009

Quantified Score

Hi-index 0.08

Visualization

Abstract

In this paper, we propose multimodal biometric image watermarking scheme through a two-stage integrity verification method using the hidden thumbnail feature vectors for safe authentication of multimodal biometrics data, face and fingerprint, respectively. It is basically blind and spread spectrum-based robust watermarking method. The proposed method enables us to detect a tampered region by controlling watermark embedding strength to meet the requirement of predefined watermark extraction threshold. The key idea is that the thumbnail feature vectors of a face image as a watermark pattern are utilized by embedding into a fingerprint image in order to verify the integrity of respective biometric data. The first stage of integrity verification for a fingerprint image is done by deciding the validity of extracted thumbnail patterns. The second stage of integrity verification for a face image is done by one-to-one matching between the thumbnail feature vectors extracted from a face image and the thumbnail one of the received face image. Experimental results showed that the proposed method has a high detection rate of the forged biometric data and guarantees the security assurance.