The Research And Implementation Of Color Image Digital Watermarking Algorithm Based On Wavelet Transform

With the development of computer multimedia technology and printing technology, people's life is more convenient while the acts of piracy and illegal tampering become more difficult to detect and certificate. Information security has become serious social and economic issues increasingly. As a branch of information hiding technology, digital watermarking technology especially the image watermarking has played a vital role in information security.There are many watermarking algorithms based on wavelet transform. The typical method is to modify the low-frequency coefficients of wavelet transform directly in order to embed the watermark. These methods have good invisibility but weak ability to resist geometric attacks and small amount of watermark. Moreover, these methods are not suitable for hardware implementation for requiring the original data. For the Inadequate of the algorithm, this paper proposed an improved color image digital watermarking algorithm basted on wavelet transform. By the adaptive threshold, the algorithm embeds the binary watermark image into the low frequency and medium frequency of Cb component. Watermarking extraction improves the robustness with geometric correction and need not original image date.Using the FPGA hardware, the paper designs a watermarking extracting system, which is suitable for hardware implementation.The main contents are as follows:(1) An introduction is made which includes the background, implementation and research status. The paper discusses the principles of digital watermarking technology, features and performance evaluation systematically, and analyzes the methods of digital watermarking typically.(2) According to the human visual system characteristics, the algorithm chose the Cb component to embed the watermark. As the experiments show, Cb component is more suitable to embed watermark than the other two components.(3)Depending on the difference between texture complexities of images, this paper use adaptive threshold to improve the traditional embedding methods that modify the wavelet coefficients directly. Compared with the method that use fixed threshold, this algorithm has better invisibility. In the low frequency and medium frequency of Cb sub-band, we define 4X4 as a block in embedding process which increases the capacity of watermark embedded. (4)Accounting to the image suffered geometric attacks, the watermarking extraction algorithm adopts the geometric correction to improve the robustness, and realizes the blind extraction that does not require the original dates. Based on the FPGA hardware platform, the paper presents a digital watermarking extraction system that is suitable for the hardware implementation. We research and design the implementation program of 5/3 lifting wavelet transform, color space conversion and watermarking extraction modules.Finally, the test of algorithm performance was made. The experimental results show that the proposed digital watermarking algorithm has not only good invisibility but also strong robustness to the noise attacks, JPEG compression, rotation, shear attacks. In a certain extent, the algorithm meets the requirements of image digital watermarking, and has some practical value.