Robust embedding of image watermark using LWT and SVD

Autor: Sushma S. Agrawal, Radhika G. Kabra
Rok vydání: 2016
Předmět:
Zdroj: 2016 International Conference on Communication and Signal Processing (ICCSP).
DOI: 10.1109/iccsp.2016.7754516
Popis: Watermarking is a method of concealing a secret image within original image, used for security purpose. This paper proposes a watermarking method grounded on Lifting Wavelet Transform and Singular Value Decomposition. In this scheme Watermark is not injected directly in Wavelet coefficients. Three level decomposition of original image is done; Singular Value Decomposition is applied to four sub-bands. Watermark image is divided into four parts. SVD is applied to each part of Watermark. Singular values of sub-bands of original image are reformed using Singular values of fragmented Watermark. Modification in all frequencies is robust to various attacks like histogram equalization, sharpening, gamma correction, Gaussian filter, re-watermarking, etc. This paper proposes considerably more imperceptible and eminent results.
Databáze: OpenAIRE