Research On Techniques Of Image Digital Watermarking And Ability Evaluation

With the development of digital technology and the Internet, various kinds of multimediadigital work, such as image, video, audio, text etc. are published on the network, whosecopyright protection and integrality authentication are quite impendent to solve. Aiming atprotecting the copyright and authenticating the integrality of digital work, the technology ofdigital watermarking is developing, by which the secret information is embed in original data,which is also called watermark. By the position in which the watermark is stored, digitalwatermarking is classified as two kinds, digital watermarking in space domain and digitalwatermarking in transform domain. Digital watermarking in space domain means to imprint thewatermark in space domain, while digital watermarking in transform domain means to imprintthe watermark in DCT domain or Wavlet domain.This paper has throughly and systemically researched on image digital watermarking andability evaluation. Main work includes three parts:1. Image watermarking technique in space domain. image watermarking technique in spacedomain means to imprint the watermark in space domain.Three of the typical methods are LSBalgorithm,Van Schyndel algorithm and Patchwork algorithm. The paper has throughly analyzedthe algorithms above, disigned and implemented the algorithms, and simulated the results. Dueto the shortage of the algorithms mentioned above, the paper puts forward an improved spacedomain image watermarking algorithm, disigned and implemented this algorithm, and simulatedthe results. Facts show this apace algorithm has strengthed the security of watermarkinginformation.2. Image watermarking technique in DCT domain. Image watermarking technique in DCTdomain means to imprint the watermark in DCT domain. Three of the typical methods are BlockDCT algorithm, DCT algorithm based on threshold and Whole DCT algorithm. The paper hasthroughly analyzed the algorithms above, disigned and implemented the algorithms, andsimulated the results. Due to the shortage of the algorithms mentioned above, the paper putsforward an improved DCT domain image watermarking algorithm, disigned and implementedthis algorithm, and simulated the results. Facts show this DCT algorithm has increased thevolume of watermarking information.3. Ability Evaluation on image watermarking. This paper introduces two variables: PSNR(Peak Signal to Noise Ratio) and sim (similarity) which describe vision quility quantificationally.We hav taken on Alility evaluation on geometry distorted attack and image strength attackrespectively, analyzed the simulated data, and compared the anti-attack alility of the two improvrd algorithms.4. On the basis of analyzing the existing Digital watermarking evaluation method, Thepaper put forward a mature digital watermarking evaluation scheme based on application scenemodels.