Speech signals protection via logo watermarking based on the time–frequency analysis

A procedure for logo watermarking of speech signals is proposed. In order to take full advantage of both time and frequency domains, we have used joint time–frequency representation of speech signal, which provides a more comprehensive information compared to each domain individually. Time–frequency...

Full description

Saved in:

Bibliographic Details
Published in:	Annales des télécommunications Vol. 63; no. 7-8; pp. 369 - 377
Main Authors:	OROVIC, Irena, ZOGOVIC, Predrag, ZARIC, Nikola, STANKOVIC, Srdjan
Format:	Journal Article
Language:	English
Published:	Paris Springer-Verlag 01-08-2008 Springer
Subjects:	Applied sciences Circuits Communications Engineering Computer Communication Networks Cryptography Detection, estimation, filtering, equalization, prediction Engineering Exact sciences and technology Information and Communication Information Systems and Communication Service Information, signal and communications theory Networks Original Paper R & D/Technology Policy Sampling, quantization Signal and communications theory Signal processing Signal, noise Signal,Image and Speech Processing Speech processing Telecommunications and information theory Speech signals Logo watermarking Time–frequency analysis Time-frequency analysis Time frequency domain method Acoustic signal Noise reduction Discrete Fourier transformation Frequency selection Vocal signal Information protection Signal quantization Speech signals, Time-frequency analysis Signal processing Robustness Sampling Pitch(acoustics) Signal analysis Digital watermarking
Online Access:	Get full text
Tags:	Add Tag No Tags, Be the first to tag this record!

Description
Summary:	A procedure for logo watermarking of speech signals is proposed. In order to take full advantage of both time and frequency domains, we have used joint time–frequency representation of speech signal, which provides a more comprehensive information compared to each domain individually. Time–frequency-based selection of the discrete Fourier transform coefficients is used for logo embedding. The proposed coefficients selection provides both the imperceptibility and robustness requirements. The binary logo is embedded by using the concept of bit-plane modification. The logo extraction procedure is provided as well. The robustness of the proposed procedure is tested under various attacks such as compression, noise addition, filtering, re-sampling, re-quantization, time stretch, pitch scaling, etc.
ISSN:	0003-4347 1958-9395
DOI:	10.1007/s12243-008-0036-3