Quantization Based Blind Watermarking Algorithm Using DWT

摘要

In this paper a Discrete Wavelet Transform based blind watermarking algorithm is proposed for embedding a grayscale watermark into a grayscale image. A new approach of choosing the watermark pixels pseudo-randomly for embedding in the mid-frequency region of the source image is proposed. A pseudo random generator, which exploits the randomness of the universal constant n, has been used. This algorithm makes use of the concept of thresholding based on the standard deviation of the subblocks in the mid-frequency region. The experimental results confirm the robustness of the watermark against noise and commonly used image-processing operations like cropping and JPEG2000 lossy compression. A novel blind watermarking technique and an efficient pseudorandom generator are major contributions of this paper.