Compression Combined Robust Watermarking Scheme using SVD Replacement Technique

Abstract

This paper proposes a novel compression combined digital image watermarking scheme based on singular value replacement technique. Image compression is achieved using Huffman encoding technique. Huffman encoding is an entropy encoding algorithm offering lossless image compression. The proposed watermarking scheme combines Integer wavelet transform (IWT) with singular value decomposition (SVD). For watermark embedding, the singular values (SV's) of high frequency (CD) band of cover images are replaced with the singular values of watermark signal. Choosing a sample image as watermark signal depends on the relation between energy of singular value of high frequency component of cover and sample image. The combination of Huffman encoding with IWT-SVD domain watermarking results in a robust watermarking scheme that provides good compression ratio with better signal quality. Experiments suggest that the watermarked and original images are perceptually similar. Also, watermarked images are robust against image processing distortions and geometrical attacks. Further, the recovered images are distortion free.