Still Image Digital Watermarking Technology

Digital watermarking refers to the information hiding technique applied in the information security field. It is an efficient means to protect intellectual property by inserting additional data into the digital media, such as images, audios, and videos. Ownership protection and content authentication are two main key problems in image watermarking technique. In this paper, several algorithms have been proposed in these two area:A robust image watermarking scheme in DCT domain is presented. By utilizing secondary DCT and the visual masking of HVS, the watermark can be embedded adaptively into the DCs (Direct Coefficient). Experimental results show that the presented watermarking scheme is transparent and robust against various signal processing such as JPEG compression, low-pass filtering, and noise adding, etc.An image authentication scheme based on edge character is presented. Combined with JPEG2000, the authentication data is generated by calculating a feature matrix in the higher sub-bands of wavelet coefficients. Then, the encrypted data would be embedded into the low-frequency sub-band. The integrity authentication is implemented by comparing the newly generated feature and the embedded data. The original image is not involved in this process. The experiment shows that the presented authentication scheme can effectively detect the event and the location precisely when the original image content is tampered viciously, and tolerate the damage produced by JPEG2000 loss compression.A robust image watermarking scheme based on IFS (Important Feature Point) is presented. Combining multi-scale Harris detector with Gaussian-Laplacian detector, the IFSs, which can survive a variety of attacks, are extracted. Besides, the local feature regions are ascertained adaptively based on image local contents and the watermark sequence is embedded into all local feature regions. Experimental results show that the presented scheme is not only robust against common signal processing such as median filtering, sharpening, noise adding, and JPEG compression, but the desynchronization attacks such as rotation, translation, scaling, row or column removal, shearing, and local random bend.