Digital Image Watermarking in Multimedia Data Compressions using Robust 3-Level Discrete Wavelet Transform and Alpha Blending

The watermark means the insertion of predefined patterns into the multimedia data, so that the reduction in quality is minimized and remains at an imperceptible level. Many digital watermarking algorithms have been proposed in special and transformational fields. Spatial techniques still have a relatively low bit capacity and are not sufficiently resistant to lossy image compression and other image processing operations. For example, a simple noise can remove the watermark. On the other hand, frequency domain based techniques can incorporate more bits for watermarks and are more robust against attacks. Certain transformations such as the discrete cosine transform (DCT) and the discrete wavelet transform (DWT) are used for watermarks in the frequency domain. In this work, we compare the DCT watermark algorithms and in particular the DWT watermark algorithms based on robustness criteria. In other words, the robustness of various transformation watermark algorithms is assessed using different attacks. One of the main reasons for considering DWT watermark algorithms is that several multimedia standards like JPEG2000 and MPEG-4 are based on DWT. These new standards have created new requirements such as progressive and low bit rate transmission and coding by regions of interest. Digital watermarking is a technique in which a watermark signal is integrated into the host image to authenticate it. The incorporation and extraction of watermarks takes place in the high frequency range of the discrete wavelet transformation (DWT), since small changes in this area are not perceived by the human eye. This watermark scheme deals with extracting the watermark information in the absence of an original image, so that the blind scheme was obtained. The peak signal-to-noise ratio (PSNR) is calculated to measure image quality. The experimental evaluation of the proposed method shows very good results with regard to robustness and transparency to various attacks such as median filtering, Gaussian noise and JPEG compression. Our project presents a digital watermark algorithm with discrete wavelet transform (DWT) based on signs of human vision. Using this technique, the watermark signal is integrated into the high frequency band of the wavelet transform domain.