Abstract
Conference Title: 2015 Global Summit on Computer & Information Technology (GSCIT) Conference Start Date: 2015, June 11 Conference End Date: 2015, June 13 Conference Location: Sousse, Tunisia Watermarking is an interesting technique for message hiding and authentication. We have applied 3- level Haar-DWT and thereafter, DCT transformation on each selected HH3 subband is computed. In extraction the same procedure is used to retrieve watermark logo image from the DCT component of middle frequency subband of selected HH3 subband. To ensure the accuracy and image quality between watermarked image and host image the Mean Absolute Error, Peak SNR, and Weighted Peak SNR is computed. The simulation result shows that the proposed algorithm has significantly higher robustness against Gaussian noise, Salt & Pepper noise and cropping attack of watermarked image.