Fulltext-Image Watermarking Using Improved DWT and HVS Based on JPEG Compression

As multimedia data and their applications are rapidly distributed on the Web, the research on watermark insertion and integrity vigorously increases. In this paper, discrete wavelet transformation dissolves an original image into the 3-levels of low, middle, and high bandwidths. Human visual system with the significant coefficients extracted from the middle obtains enhanced image quality and robustness using watermark insertion. In the end, experimental results show that various attacks cannot destroy pre-inserted watermark and recovered image is of good quality by measuring PSNR.