Open AccessOptimal Colour Image Watermarking Using Neural Networks and Multiobjective Memetic Optimization
Author(s) -
Hieu V. Dang,
Witold Kinsner
Publication year - 2022
Publication title -
international journal of neural networks and advanced applications
Language(s) - English
Resource type - Journals
ISSN - 2313-0563
DOI - 10.46300/91016.2022.9.5
Subject(s) - digital watermarking , watermark , robustness (evolution) , embedding , artificial intelligence , computer science , wavelet , image (mathematics) , mathematics , mathematical optimization , pattern recognition (psychology) , algorithm , biochemistry , chemistry , gene
This paper deals with the problem of robust and perceptual logo watermarking for colour images. In particular, we investigate trade-off factors in designing efficient watermarking techniques to maximize the quality of watermarked images and the robustness of watermark. With the fixed size of a logo watermark, there is a conflict between these two objectives, thus a multiobjective optimization problem is introduced. We propose to use a hybrid between general regression neural networks (GRNNs) and multiobjective memetic algorithms (MOMA) to solve this challenging problem. Specifically, a GRNN is used for efficient watermark embedding and extraction in the wavelet domain. Optimal watermark 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.