Header menu link for other important links
X
Selective image watermarking through normalized principal components
N. Sangeetha, ,
Published in Institute of Electrical and Electronics Engineers Inc.
2020
Abstract
Image watermarking is one among the preferred solutions to deal with copyright and ownership problems. This paper suggests a watermarking method in the selective blocks of the cover image using covariance analysis. To accomplish watermarking, the cover image blocks and the watermark (WM) scrambles are selected with the same spatial resolution. The Pearson correlation is calculated between each block of the cover image and the WM scrambles obtained using Arnold transform. Five pairs of cover image block and WM scramble with maximum correlation are selected for watermark embedding process. During the extraction steps, five different scrambles of the watermark are extracted from five embedded blocks of the cover image. Watermark embedding is implemented using normalized principal components based discrete wavelet transform-singular value decomposition watermarking scheme and the Arnold transform is used to obtain the scrambles of the WM. The performance of the proposed watermarking technique is analysed by the metrics evaluated on embedding and extraction. © 2020 IEEE.