Based On Wavelet Transform Image Digital Watermarking Technology Research

With the development at full speed of the multimedia technology and network technology, people obtaining to digital products is extremely fast and convenient, but how are copyright of the original creators of these digital products and economic benefits protected? Are the digital products safe and believing? Centre on this question, it is near and putting forward a kind of new effective digital products copyright protection and the new technology of security maintenance - the technology of digital watermark (Digital Watermark ) in the world in a few years.Digital watermark technology is a kind of information safe practice, it includes very extensive research field such as the hide writing, encryption technology, communication theory and digital signal processing etc. latently, the watermark is to the digital target who is protected (Such as the image, video, audio frequency, etc.) the one that imbedded can prove the information that the copyright is belonged to or can follow the tort and can be proposed and got the copyright which prove the carrier out when being necessary , what the watermark can't be separated with the carrier but also can resist general treatment to operate and attack.This dissertation has carried on overall research to digital watermark technology.First of all, have introduced the essential feature , principle and present domestic and international research current situation of the digital watermark, essential feature and classification of including the digital watermark, have summed up the basic frame of the digital watermark system, generality has introduced the common digital watermark algorithm at present.Secondly, brief to introduce Wavelet Transform definition. Mallat algorithm, two-dimensional wavelet decomposability of image and orthogonal wavelet decomposability. And the zero tree structure of image wavelet transform.Then, have proposed a kind of scheme of digital watermark based on content characteristic of the image. The watermark adopts the grey level picture directly, fully consider while imbedding content characteristic and human eye vision of the picture shelter (HVS) the characteristic, imbed the watermark in the weight of lowintermediate frequency. Abstracting the edge information of the picture first, carry on coefficient to revise to it; The coefficient characteristic after varying according to the small wave, there is the intensity of imbedding the watermark of adjustment that is chosen, make the robust getting excellent and inobservance raise.Afterwards, proposed a kind of adaptive digital watermark algorithm on the basis of the fuzzy cluster. Watermark picture adopt two-value picture, is it put to watermark messy to go on first, and then to put messy watermark after is it modulate as the watermark that is imbedded to go on with two-value false random array. In embed course, carry on wavelet transform first, choose to embed the frequency band of the watermark, to selected bringing into the conduct FCM algorithm frequently, carrying on the fuzzy cluster. Then the right embedding watermark of since the result to the cluster. And has verified the validity of the algorithm with the artificial experiment.And then to carry on the research that the blind watermark measures.My work still includes:Proposed one image fusion algorithm based on feature wavelet transform and high-pass filter. It is a kind of important enhancement image information method that the picture is fused. This dissertation has put forward the fusion method based on many-level decompose algorithm, considering edge characteristic and the picture vary coefficient dependence after wavelet transform, carrying on the high pass and straining waves to the weight of low frequency, the weight of detail of getting the weight of low frequency is merged; It draws method that edge combine regional energy of picture merge to adopt in high-frequency area. The experimental result proves this algorithm can get very good integration result.A self-adaptive image fusion algorithm based on region characteristic of wavelet transform coefficients is presented, and comprehensively considered the features of image's edge and the coefficient correlation after wavelet transformation. In low frequency area, the images are fused on weighted fusion: While in high frequency area, the image's edge is abstracted first, then the images are fused by the approach based on the dynamic weighting of feature. The experimental result has proved the validity of this algorithmFinally, summarized my result of study and looked forward to the further development of the digital watermark.