The Research On Fractal Feature Extraction Algorithm Of Wood Cell Image

With the development of modern wood industry, wood as the basic form of wood industrial product, different species of wood is frequently used in people's lives.However, they have a lot of differences between the timber species, it is difficult to recognize them. Therefore, we should solve the problem how to recognize wood accurately, by using auxiliary equipments in the precondition that do not destroy, product. And the problem becomes the research direction at present.Fractal dimension of wood texture cell image can represent a lot of texture features, and it has important wood texture information. So we can conclude that it is an important characteristic parameter of timber species. The major purpose of this dissertation is to study wood texture images. Taken for the image processing technology, we have distilled the fractal dimension of wood texture images in wood section which are recorded by the video.In order to complete it, we use improved box-counting approach to compute the fractal dimension of wood texture cell image. In terms of the theory analysis and experiments, it is proved that the method can extract the characteristics of wood texture,and it can overcome the influence of the proportion size.It is an important method of the wood texture feature extraction, and it is worth our deep research.For studying the texture feature of timber species further, this dissertation proposed a new fractal processing method that is much better than the traditional algorithm.Because the traditional algorithm has.a lot of limitation, we divide the wood texture cell image into different regions, and then extract the fractal dimension of the regions.The experiments prove that the method can extract the characteristics of wood texture. Moreover, it can better reflect the texture features of timber species in different regions of wood texture cell image. Just because the traditional algorithm is for the whole image. The method is good for the study of timber species, and it is an important method.This paper is a part of "Research on numerical control equipment of wood recognition" which is supported by tackle key problem item of Heilongjiang Province.The dissertation makes some the quantitative research to the wood texture feature of timber species in the microscopic domain. In the paper the system software is designed by Visual C++6.0 programming language and the modular designing method. In addition, the extracted fractal texture parameter of timber species is an important digital parameter, especially identifying timber species. And it provides a large amount of data for the identification of wood species. The study in this paper will provide well theoretical and applied foundation for the wood recognition.