Color-blind Watermarking Method Based on Hamming Code, Eigenvalue Decomposition and Image Correction
A technology of eigenvalue decomposition and image correction, applied in image data processing, image data processing, instruments, etc., can solve problems such as image noise interference and increased difficulty of blind watermark detection
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Publication Date
- 2022-04-05
Smart Images

Figure 1 
Figure 2 
Figure 3
Abstract
Description
technical field
[0001] The invention belongs to the field of network space security and relates to the copyright protection of color digital images as watermarks. Background technique
[0002] With the rapid development of Internet technology, the amount of information on the network is increasing rapidly, and color digital images, as carriers of large amounts of information, are increasingly entering people's lives, bringing great convenience to people. At the same time, the network Some illegal infringements such as distortion, tampering, and plagiarism on the Internet have also followed. In addition, the carrier image with a large amount of information is more likely to be disturbed by noise. The robustness of the watermark algorithm is facing severe challenges. The copyright of color digital images The problem of protection has been widely concerned by people; at the same time, the damage to the watermark contained in the carrier by image geometric processing makes it mo...
Examples
Embodiment Construction
[0071] The purpose of the present invention is to provide a color-blind watermarking method based on Hamming code, eigenvalue decomposition and image correction, which is characterized in that it is realized through four processes, specifically including a preprocessing process, a watermark embedding process, an image correction process and a watermarking process. Extraction process:
[0072] Process 1: Preprocessing process, the specific steps are described as follows:
[0073] Step 1: Preprocessing of the color host image: Perform dimensionality reduction processing on the 3D color host image H with a size of 512×512×3 to obtain three 2D layered host images H of red, green and blue p , and divide the pixels in each layer into 2×2 non-overlapping pixel blocks, p=1, 2, 3 respectively represent three layers of red, green and blue;
[0074] Step 2: Preprocessing of the color watermark image: Perform dimensionality reduction processing on the 3D color watermark image W with a si...