Open AccessIntelligent High Payload Audio Watermarking Algorithm Using Colour Image in DWT-SVD Domain
Author(s) -
A. R. Elshazly,
Mohamed E. Nasr,
Marwa Fouad,
Fathi. E. Abd-El-Samie
Publication year - 2021
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/2128/1/012019
Subject(s) - digital watermarking , watermark , payload (computing) , discrete wavelet transform , computer science , artificial intelligence , audio signal , computer vision , algorithm , singular value decomposition , grayscale , wavelet , wavelet transform , speech recognition , embedding , pixel , image (mathematics) , speech coding , computer network , network packet
Copyright protection and ownership verification of digital audio tracks have become increasingly important to be enabled by digital watermarking techniques. A novel high payload intelligent audio watermarking scheme with RGB color watermark image is proposed in this paper. The color watermark image is encrypted using Arnold chaotic map and passed through an adaptive scaling filter to scale the image to match the required payload. The encoding process is performed on the scaled encrypted version of the watermark image. A portion of the audio signal is used to embed a synchronization code and the other one is decomposed into short frames. These frames are processed with a two-level discrete wavelet transform (DWT), followed by a singular value decomposition (SVD) process on the approximation coefficients. The encoded watermark is inserted into the diagonal matrix using quantization index modulation (QIM). The inverse process of SVD and DWT is applied to obtain the marked audio signal. Blind extraction of the hidden information from the marked audio signal is performed in the reverse order of the embedding process. Experiments show that security, high payload, transparency and imperceptibility of the algorithm are satisfied. The robustness against several kinds of audio signal processing attacks is shown. Performance evaluation tests with SNR, BER, and FSIM are conducted.