A Study of Digital Image Watermarking for JPEG and PNG Images using Discrete Wavelet Transform

Abstract

Objectives: To emphasis on digital image watermarking techniques and its applications using Discrete Wavelet Transform (DWT). Methods: We propose two watermarking algorithms for embedding and extraction process and use two images as cover image and watermark image of same format as input image. We have considered two image formats JPEG (Joint Photographic Experts Group) and PNG (Portable Network Graphics) of different resolutions for analyzing the performance of the proposed algorithm. Findings: The proposed research work is compared with the existing works on the basis of compression percentage, MSE (Mean Square Error), PSNR (Peak Signal to Noise Ratio) and Normalized Correlation (NC) values and the result has been carried out using DWT in Matlab 7.12. All the existing watermarking techniques referred in this paper have some limitations at somewhere. The PSNR values of all the existing research works is not more than 50 dB for the entire traditional image processing images. But in this proposed research work, the PSNR value is more than 66 dB for all the twenty test images we have taken. Higher the PSNR value means the quality of the watermark image is better. If the PSNR value is good then the MSE value will automatically be good because one is inversely proportional to other. Improvements: In the previous research works, common values of PSNR were between 30 dB to 50 dB. If the PSNR value of the watermark image is more than 50 dB then it will be tough to detect the difference between the cover image and the watermark image by the common Human Visibility System (HVS). The NC value 1 means that the correlation between the extracted watermark image and the original watermark image is almost similar in a particular point.