Digital speech watermarking to enhance the security using speech as a biometric for person authentication

Revathi, A.; Sasikaladevi, N.; Jeyalakshmi, C.

doi:10.1007/s10772-018-09563-9

Cited by 7 publications

(1 citation statement)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the second set, biometric data is embedded into natural images to protect intellectual property rights and copyright. Various biometric data such as speech signal [10,11], fingerprint [12][13][14], face [15,16], iris [17][18][19][20], and signature [21] are used as watermark for this purpose. Authors in [18] and [19] embedded iris code into cover image using Discrete Wavelet Transform (DWT)-Singular Value Decomposition (SVD)-based and Discrete Cosine Transform (DCT)-SVD-based hybrid watermarking schemes, respectively.…”

Section: Introductionmentioning

confidence: 99%

Biometric Watermarking Schemes Based on QR Decomposition and Schur Decompositon in RIDWT Domain

Altay

Ulutaş

2023

Preprint

View full text Add to dashboard Cite

In this paper, two new methods are proposed to hide a biometric-based digital watermark generated from an iris image into a color image to resolve ownership conflicts. One of them is the QR decomposition-based approach in the redistributed invariant discrete wavelet transform (RIDWT) domain, and the other is the Schur decomposition-based approach in the RIDWT domain. After the Y channel of the color image in YCbCr space is decomposed by RIDWT, the lowest frequency sub-band is divided into non-overlapping blocks. Blocks with a lower standard deviation are selected to embed each bit of the watermark created from the iris image into a block. Watermarking of selected blocks is performed with QR decomposition in the first approach and Schur decomposition in the second approach. The firefly algorithm is used to balance the robustness and perceptual transparency in hiding the watermark. Experimental results are evaluated in terms of robustness and imperceptibility to prove the effectiveness of the proposed systems. In addition, validation performances are analyzed based on false rejection rate, false acceptance rate and equal error rate. The results show that both methods have high perceptual quality and robustness, and also a very low error rate in ownership verification.

show abstract