Image Watermarking Based On Digital Holography And DCT-SVD Algorithm Research

The development of computer and communication technology has brought convenience to the communication of data and information such as image, video, etc., and they also increase the risk of unauthorized information copying and the theft. The protection of intellectual property rights become the widely discussed problem. The one of the effective ways to solve this problem is watermark technology. Watermark technology embeds the secret information that has the marked effect into digital image products. The watermark information can be extracted to verify the integrity of the product and legitimacy by specific means.(1) Because of fast operation, good security and high robustness, the security technology based on optical theory has been widely used in digital watermarking. Existing additive based on digital holography and DCT transform embedding algorithm, the algorithm has simple structure, security and robustness are not ideal. The existing algorithms based on digital holography and DC T(DCT for Discrete ’Cosine Transform) transform has the simple structure, but the security and robustness is poor. Towards the shortcomings of the existed in the existing algorithms, this paper puts forward the algorithm based on digital holography and DCT transformation of multiplicative embedding. This algorithm applies law of multiplicative to embed the watermark image into the original image. Therefore the algorithm increases the difficulty of the embedding algorithm, but accordingly increases the security of the algorithm and improves the robustness of the watermark. Firstly, watermark image is Fresnel transformed to get the digital hologram. Then the digital hologram is Arnold scrambling transformed to get the embedded watermark. The original image is transformed by DCT transform. Finally the embedded watermark is embedded into the original image by multiplicative embedding rule. The wavelength and diffraction distance are used as the key and have higher security. Experiments prove that this algorithm has higher robustness, but the ability to resist geometric attacks remains to be improved.(2) In order to solve the algorithm of deficiencies in the third chapter, combined with the stability of the Singular Value Decomposition, this paper presents a watermarking algorithm based on DC T and SVD(SVD for Singular Value Decomposition). The SVD decomposition can better describe the characteristics of the image, improve the confidentiality and robustness of the watermark. Experiments show that the algorithm has a higher robustness against cutting operation, noise pollution, image compression and rotation attacks.