| Multifractality is a new branch of nonlinear science,as a core branch of fractal geometry for researching irregular geometrical figure,it has been used in the other fields of science more and more widely.Especially,its potential value in theory and application of the image processing,makes the application of multifractal geometry theory in image processing become a research subiect of many researchers.Multifractal is an effective mathematical tool for descripting the structure of the singularity.Multifractal is used to describe the odd probability distribution from which can not be described completely by one characteristic scale exponent that is determined by the fractal dimension.It describes different characteristics in different level in a growing process of fractal geometry body which has scale property.Every level can be indicated in different characteristic parameters, which compose one set or spectral function.In this paper,proposed the method which use the eight multifractal spectral characteristics to recognize image after in-depth study of the multifractal theory,feature extraction algorithm and its application.The eight multifractal spectral characteristics areαmin,αmax,fmin,fmax,spectral iwidth,spectral dispersion,αd and asymmetry degree.Through the simulation program,recognize 200 images(Three types),and the recognition rate reached 80%.With using multifractal approach,it might shed some light on the theory of image recognition.Simulation results show that the method of mutifracal is well validate for recognizing objects and the eigenvalue extracted by this method have a more effect on the identification of images.The achievement of our research results enriches the approaches to the image classification and recognition technology and its applications,and has a certain theoretical and practical importance.This thesis provides useful methods and approaches for the research and the development of image classification and recongnition technology and its applications. |
PDF Full Text Request