Robust biometric image watermarking for fingerprint and face template protection

Vatsa, Mayank; Singh, Richa; Noore, Afzel; Houck, Max M.; Morris, Keith B.

doi:10.1587/elex.3.23

Cited by 86 publications

(32 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Vatsa et al [14] combined discrete wavelet transform (DWT) and the least significant bit (LSB) methods for watermarking biometrics. Watermarking technique based on the DWT method is robust to frequency attacks but vulnerable to geometric attacks.…”

Section: Related Workmentioning

confidence: 99%

Digital Image Watermarking Techniques and Biometrics Data Security: A Review

Adamović¹

2017

Proceedings of the International Scientific Conference - Sinteza 2017

View full text Add to dashboard Cite

Abstract:In this study, we review digital watermarking techniques used for protection of biometrical data in authentication services based on biometry. In addition to the development of protection biometrical templates, there has been a breakthrough in additional privacy protection within biometrical systems that perform authentication without preserving biometry in their databases. We begin with theoretical foundations of digital watermark and biometry and then review current research advances in this area, which combine the two technologies and thus create a new ground for exploration better known as biometrical watermarking.

show abstract

Section: Related Workmentioning

confidence: 99%

Digital Image Watermarking Techniques and Biometrics Data Security: A Review

Adamović¹

2017

Proceedings of the International Scientific Conference - Sinteza 2017

View full text Add to dashboard Cite

show abstract

“…In this proposed technique, Mel frequency cepstral coefficients (MFCC) is embed into red and blue channel of color face image for improve robustness and security of multimodal biometric system. Author in [11] proposed biometric watermarking technique based on LSB and DWT for embedding face features into fingerprint image for multimodal biometric system. Author in [12] proposed biometric technique based on multi resolution DWT and support vector machine (SVM) for embedding face image into fingerprint and claimed that this technique improved 10 % of face recognition under different attack.…”

Section: Literature Review Of Related Workmentioning

confidence: 99%

Compressive Sensing Based Multiple Watermarking Technique for Biometric Template Protection

Thanki¹,

Borisagar²

2014

IJIGSP

View full text Add to dashboard Cite

Abstract-Biometric authentication system is having several security issues. Two security issues are template protection at system database and at communication channel between system database and matcher subsystem of biometric system. In this paper, two level watermarking technique based on CS Theory framework in wavelet domain is proposed for security and authentication of biometric template at these two vulnerable points. In the proposed technique, generate sparse measurement information of fingerprint and iris biometric template using CS theory framework. This sparse measurement information is used as secure watermark information which is embedding into a face image of same individual for generation of multimodal biometric template. Sparse watermark information is computed using Discrete Wavelet transform (DWT) and random seed. The proposed watermarking technique not only provide protection to biometric templates, it also gives computational security against spoofing attack because of it is difficult for imposter to get three secure biometric template information where two encoded biometric template is embed in term of sparse measurement information into third biometric template. Similarity value between original watermark image and reconstructed watermark image is the measuring factor for identification and authentication. The experimental results show that the technique is robust against various attacks.

show abstract

“…Their face recognition accuracy is reported above 80% and was improved to 97% with the aid of SVM. In the subsequent work [21], they proposed a combined DWT and LSB (Least Significant Bit) based biometric watermarking algorithm that securely embeds a face template in a fingerprint image. The proposed algorithm is against geometric and frequency attacks and protects the integrity of both the face template and the fingerprint image.…”

Section: Pankanti and Yeungmentioning

confidence: 99%