Traffic Image Classification Method Based on Fractal Dimension

Author

Wen-lun, Cao ; Shi Zhong-ke ; Feng Jian -hu

Author_Institution

Sch. of Autom., Northwestern Polytech. Univ., Xi´´an

Volume

2

fYear

2006

fDate

17-19 July 2006

Firstpage

903

Lastpage

907

Abstract

An image classification method based on fractal dimension was presented. Moreover, the fractal dimension of small complex traffic image was studied. The result of the experiment show fractal dimension can be a good parameter in image classified and analysis. The classified result based on fractal dimension is the same as the result of classified according to person´s subjective recept. To organize in a sequence of texture-image, texture-verge-image, texture-verge-smooth-image, verge-smooth-image and smooth-image, the corresponding fractal dimension respectively is 2-2.1,2.1-2.35,2.35-2.45,2.45-2.5,2.5-3.0 and 2-2.2,2.2-2.45,2.45-2.55,2.55-2.6,2.6-3.0, when the box´s side is 2 and 4. As for traffic image, the fractal dimension respectively is 2.4-3.0 and 2.8-3.0, when box´s side is 2 and 4

Keywords

fractals; image classification; image sequences; image texture; traffic engineering computing; fractal dimension; texture-verge-smooth-image sequence; traffic image classification; traffic image processing; Arithmetic; Clouds; Fractals; Image analysis; Image classification; Image coding; Image processing; Pixel; Sea surface; Space technology; Fractal Dimension; Image Classification; Traffic Image Processing;

fLanguage

English

Publisher

ieee

Conference_Titel

Cognitive Informatics, 2006. ICCI 2006. 5th IEEE International Conference on

Conference_Location

Beijing

Print_ISBN

1-4244-0475-4

Type

conf

DOI

10.1109/COGINF.2006.365612

Filename

4216530