Using Wavelets and Independent Component Analysis for Quantization Index Modulation Watermarking

Jinhua Liu,Kun She,William Zhu

DOI: https://doi.org/10.1007/978-3-642-02962-2_70

2009-01-01

Abstract:In this paper, a watermarking algorithm based on Multiband Wavelet and Independent Component Analysis for Quantization Index Modulation Watermarking was proposed. Unlike the watermark algorithms based on traditional wavelet analysis, where the watermark bits are embedded directly on the middle-frequency wavelet coefficients in the sub image. When we extract the watermark, we applied Independent Component Analysis to extract true watermark. Firstly, we used Primary Component Analysis to preprocess, and then used Independent Component Analysis to extract the watermark. Experimental results showed the proposed algorithm has good robustness against JPEG compression, White Gaussian Noise etc.