| A new fractal subject set up Mandelbrot in 1975 with his book一fractal geometry.It raised the enthusiasm of the study Fractal and their characters for especial purpose.At the same time,many leaders of the other subjects to do their scientific research,such as physics, chemist,biology.The basic knowledge of Fractal is related to the popular subject-Chaos.In mathematics term,Fractal and Chaos are the impartibility.Cellular Automata was used to simulate the growth of the plant and the molecule's motion.Considerable interest has been shown in the past decade in the use of the CA for information processing.The main challenge is to associate a given phenomenon with the evolving automata field.It also can be used in the image processing field.Firstly,this paper introduces the history,the concepts,the basic theory and the development of CA,according to S.Wolfram's class of CA,we realized dynamic simulation of all elementary cellular automata(ECA) by using Matlab and character the dynamic behavior of stable pattern,period pattern,chaos pattern and complex pattern,including 32 types ECA.In addition,this paper give the evolvement and algorithm of odd2even regular CA,game of life and ant's regular CA,impolder the visual model of CA,this show the advantage of CA in simulating many complex phenomena.Secondly,this paper give some application of CA in image processing,including edge detection,smoothing,filtering,noise reduction,etc,and we give the CA model,algorithm and local transform rule,this show advantage of CA in image processing by comparing with tradition algorithm in some examples.But these research and application are fewer,so it's necessary to enhancing more research.Finally,this paper introduce simple conception of cellular neural network(CNN),the relation of CA and CNN,the application of CNN in image processing,point out CA is a perfect tool in designing CNN,their combine sufficiency exert the advantage in parallel computation and have wide application foreground. |
PDF Full Text Request