Algorithm Research Of Image Processing Based On Wavelet And Fractal

In the processing of image analysis,it will result in the problem of incomplete,inexact and non- structured of the imformation because of the complexity of the image,it can not solve the problem with traditional method,so we should use new and advanced mthod to make up for the deficiency of the traditional method。These new methods include wavelet analysis,fractal theory etc。these theories have their own characteristics,these theories can also solve problems in some extent,but if we use the correlation of them,we can solve the above problems better when the wavelet and fractal are integrated。Wavelet analysis is internationally recognized up to the minute tool for analyzing time-frequency,And it was widely used in image and signal。The appearance of fractal provides a new perspective for integrated and local relationship of things,and fractal dimension is the main characteristics of it。wavelet has predestined relationship with fractal after the apperance of them,wavelet is studying on the mthod of local's dependence on whole,and fractal is studying on the local signal to determinate the whole feature of signal,they both concern on the local details of objects。In this paper,we mainly discuss the application in image processing based on the wavelet and fractal。The fundamental theories of wavelet and fractal are discussed in detail, The fast algorithm of discrete dyadic wavelet transform and the calculation of fractal parameters are given.Finally,an analysis is made on the properties of the wavelet basis,we also study the searching algorithm of the best wavelet basis and the selection of the best wavelet basis based on fractal entropy。The application of fractal in image de-noising of wavelet is introduced in detail, studying the relationship between the fractal constant and the vanishing moments of wavelet,and we can select the best de-noising wavelet basis based on fractal constant.The analysis of the choice of some parameters in the process of de-noising is made in detail,some key problems on de-noising method based on wavelet threshold are discussed in detail,and a improved thresholding mehod is proposed,and the simulation testing has proved the effectiveness of the improved scheme。The theories of second generation wavelet and dual-tree complex wavelet are discuessed, we have also proposed two noval de-noising method based on seal-adaptive second wavelet and dual-tree complex wavelet。Finally the testings haved proved the effectiveness of thoes two noval method,and the experiments show that those two noval proposed algorithm is efficient in image de-noising,so they both have good application value in image de-noising。The application of fractal in image edge detection of wavelet is introduced in detail,studying on the selection of the best edge wavelet basis based on fractal。The traditional methods of edge detection are based on one-order derivative's maximum, or two-order derivative's zero-crossing,and they are very sensitive to noise。In this paper,we choose the best edge wavelet basis by using of the fractal constant of different images。It was proved that this choosing edge wavelet method is quite effective,the effect of edge detection would keep edge positions very well。A adaptive threshold edge detection algorithm of wavelet is put forward in this paper, and the testings haved proved the effectiveness of this method,the edge effect is better than traditional method of wavelet based on modulus maximum,the experiment also proved that the method proposed have very good application value in image edge detection。The application of wavelet in image fusion is studied。Different wavelet basis algorithm in image fusion and the influence of fusion effect when using different decomposing levels are introduced in detail。In this paper,we present a scheme of multi-resolution for two different images using wavelet transform。The two different images were decomposed to some sub-band images in different frequence,we can process the different cofficiences in different frequence and could get new wavelet cofficiences,Then the wavelet cofficiences matrix was transformed inversely,Finally we could get an fused image。At last,the experiment verified the effectiveness of this scheme and the fusion effect was better,it also show that this method proposed have very good application value in image fusion。...