A Novel Multiobjective Memetic Algorithm Based Data Hiding for Colour Image Protections
Hieu Van Dang · 2013
Data hiding is the technique of embedding information (such as a watermark) into a carrier signal (video, image, audio, text) such that the information can be detected and extracted for steganography applications such as covert communication, and watermarking applications such as copyright protection, content authentication, signature verification, content identification, fingerprinting, media forensics, copy control, and broadcast monitoring [PeAK99], [MoKo05], [WuLi03]. The important technical issues for the data hiding techniques are transparency, robustness, capacity, security, and detectability [MaDD04], [MoKo05]. These issues are also requirements or objectives for data hiding systems [DaKi12]. However, there is a tradeoff between these objectives that makes the data hiding technique be considered as a difficult multiobjective optimization problem [DaKi13]. This paper focuses on solving the multiobjective optimization problem for a robust and perceptual image data hiding. In this paper, we present a novel multiobjective memetic algorithm (MOMA) based data hiding technique for colour image protections. First, the RGB colour image is converted to YCrCb colour image, and then the luminance component Y is decomposed by a discrete wavelet transform (DWT). The watermark bits are embedded into selected wavelet coefficients by training a general regression neural networks (GRNN). At the decoder, the trained GRNN is used to recover the watermark from the watermarked image. Optimal embedding factors and the smooth parameter of the GRNN are searched by a MOMA for optimally embedding watermark bits into wavelet coefficients. The experimental results show that the proposed approach achieves robustness and imperceptibility in watermarking [DaKi13].