|Table of Contents|

Fractal Dimension and GGCM Meteorology Cloud Pictures Recognition(PDF)

《南京理工大学学报》(自然科学版)[ISSN:1005-9830/CN:32-1397/N]

Issue:
1999年04期
Page:
289-292
Research Field:
Publishing date:

Info

Title:
Fractal Dimension and GGCM Meteorology Cloud Pictures Recognition
Author(s):
XiaDeshen JinSheng WangJian
Department of Computer Science and Technology,NUST,Nanjing 210094
Keywords:
satellite cloud pictures pat tern recog nit ion image processing gray gradient cooccurence matrix tex ture analysis
PACS:
TP391.41
DOI:
-
Abstract:
The image fractal dimension feature describes the complex ity and the roughness of texture. The image g ray gradient checks up the skip part of image g ray. Combining the grad-i ent informat ion w ith gray level co-occurence matrix, it w ill be bet ter to include the information of tex ture elements and its arrangement . Combined the fractal dimension w ith the 2-order stat istical features of Gray Gradient Co-occurence Mat rix ( GGCM) of 4 kinds of basic clouds pictures ( Cirrus, Cumulus, Cumulonimbus and Layer Cloud) , the clouds are classif ied and recognized. T he samples experiments show ed that with the combination of the fractal dimension and the 2-order statistical features of GGCM, the meteorological cloud pictures can be recog nized effect ively.

References:

1 Portille-Garcia J, T rueba- Santander I, de Mig ue-l Vela G. Efficient multispectral texture segmentation using multivariate statistics. IEEE Pr oceedings, Vision, Imag e, and Signal Processing , 1998,145( 5) : 357~ 364
2 Bouman C A. Multi resolut ion segmentatio n of textured images. IEEE PAMI, 1991, 6: 99~ 113
3 Anderson T W. An I ntroduction to Multiv ar iate Statistical Analysis. New Yorks: Jo hn Wiley &Sons, 1984
4 Fan Z, Cohen F S. Textured image segmentat ion as a multiple hypo thesis test. IEEE Transaction,Circuits System, 1998, cs- 35: 691~ 702

Memo

Memo:
-
Last Update: 2013-03-29