📝 Original Info
- Title: An Efficient Watermarking Algorithm to Improve Payload and Robustness without Affecting Image Perceptual Quality
- ArXiv ID: 1004.4467
- Date: 2010-04-26
- Authors: Er. Deepak Aggarwal, Er. Sandeep Kaur, Er. Anantdeep
📝 Abstract
Capacity, Robustness, & Perceptual quality of watermark data are very important issues to be considered. A lot of research is going on to increase these parameters for watermarking of the digital images, as there is always a tradeoff among them. . In this paper an efficient watermarking algorithm to improve payload and robustness without affecting perceptual quality of image data based on DWT is discussed. The aim of the paper is to employ the nested watermarks in wavelet domain which increases the capacity and ultimately the robustness against attacks and selection of different scaling factor values for LL & HH bands and during embedding not to create the visible artifacts in the original image and therefore the original and watermarked image is similar.
💡 Deep Analysis
This research explores the key findings and methodology presented in the paper: An Efficient Watermarking Algorithm to Improve Payload and Robustness without Affecting Image Perceptual Quality.
Capacity, Robustness, & Perceptual quality of watermark data are very important issues to be considered. A lot of research is going on to increase these parameters for watermarking of the digital images, as there is always a tradeoff among them. . In this paper an efficient watermarking algorithm to improve payload and robustness without affecting perceptual quality of image data based on DWT is discussed. The aim of the paper is to employ the nested watermarks in wavelet domain which increases the capacity and ultimately the robustness against attacks and selection of different scaling factor values for LL & HH bands and during embedding not to create the visible artifacts in the original image and therefore the original and watermarked image is similar.
📄 Full Content
To increase payload the one watermark is embedded in the other i.e nested watermark is used. Two different visual watermarks are used, nested and thus the resulting watermark is embedded in lower (LL) & high frequency (HH) bands based on the optimal selection of different scaling factors for each band. The DWT coefficients in lower frequency band are larger as compared to the higher frequency bands. Therefore the value of scaling factor is kept larger in lower frequency band and lower in high frequency band. The scaling factors are chosen such invisibility and quality of extracted watermark is maintained. Visual quality of extracted watermarks is measured by the Similarity Ratio (SR) between compared images.In this paper we are giving a new image watermarking method. The embedding and extraction of watermark is based on discrete wavelet transform. Matlab 6 [28] is used to implement all the coding related to digital image processing .It is a non-blind watermarking method.
Inputs: Primary watermark, secondary watermark image. Steps:
Read the primary visual watermark image.
Decompose the primary watermark image into cap1, chp1, cvp1, cdp1 bands using daubachesis filter. PSNR dB = 10 log 10 ( MAX 2 )
M S E
The subjective evaluation of extracted watermark is based on Similarity Ratio SR.
Where S denotes the number of matching pixel values in compared images, and D denotes the number of different pixel values in compared images. PSNR2 -PSNR of gray scale cover image after embedding Nested watermark.
MSE2 -MSE of gray scale cover image after embedding Nested watermark.
In our Watermarking technique the embedding capacity is more than normal Watermarking because here watermark nesting is used. . In all attacked cases the SR value is more than 0.7. (5) Robustness is improved in all the attacked watermarked images as PSNR shows a good value against the original watermarked image. (6) The scaling factors are chosen such as invisibility and quality of extracted watermark is maintained.
Most of the research is going in watermarking the text, audio, video data for copyright protection and authentication of electronic documents and media. Watermarking is the necessity for digital images as these are available at Internet without any cost, which needs to be protected. The watermarking technique that is given in this paper can be further improved to increase the hiding capacity and Robustness for RGB and the Indexed images.
- Read the secondary visual watermark image. α is a scaling value as set to 0.04. 4) Modify the diagonal DWT coefficient by adding the nested watermark image as in the equation a. Cd1(i,j)=cd1(i,j)+( αnested watermark) 5) Where Cd1 & cd1 are the modified & original diagonal coefficients of cover image and α is set to value 0.01. 6) 7) Calculate the visual quality of extracted watermark from attacked image by the Similarity Ratio (SR) between compared images. SR= S/(S+D). Outputs: Extracted Watermarks from approximation & diagonal coefficients of watermarked cover image & Attacked cover image 3 EXPERIMENTAL RESULTS In our experimental results Lena cover image of 512512 size, Primary and Secondary watermark of 64*64 size is used .Cover image is subjected to the watermark embedding and extraction. We measure the quality of water-marked images in terms of PSNR (
📸 Image Gallery
Reference
This content is AI-processed based on open access ArXiv data.
