t
h
t2
Zhaobao Zheng
5 CONCLUSION
In this paper we have shown how to apply genetic algorithm in deciding MRF parameters and use these parameters to
classify texture images. Experiment results show the effectiveness and practicality to apply genetic algorithm in texture
classification. Further work needs to be carried out in combining the MRF parameters decided by GA with other texture
features to improve the recognition rate further.
ACKNOWLEDGEMENTS
The authors would like to thank the project(No.49871067) of the Nation Nature Science Foundation for supporting.
REFERENCES
B.Bhanu, S.Lee and J.Ming, 1989. Adaptive image segmentation using a genetic algorithm, Image Understanding
Workshop,pp.1043-1055.
D.E. Goldberg, 1989. Genetic Algorithm in Search, Optimization and Machining Learning, Addison-Wesley
Co., Massachussetts .
G.Roth and M.D.Levine,1991. A genetic algorithm for primitive extraction, Proc. 4th Intl Conf. Genetic Algorithm, San
Diego.,pp.487-497.
Hong Zheng and Zhaobao Zheng, 1997. The knowledge acquirement for image understanding based on MRF, Journal
of Wuhan Technical University of Surveying and Mapping, Vol.22,No.4,pp314-317.
Kashhyep R.L .Chellappa R. And Ahuja N., 1981. Decision rules for choice of neighbors in random field models of
images, Computer Graphics and Image Processing, Vol.15,pp.301~308.
Zhaobao Zheng and Guilan Huang,1996. The least square method for aerial image texture classification and its problem
analysis, Journal of Surveying and Mapping, Vol.25,No.2, pp.121~126.
Zhaobao Zheng and Yueqing Zhou, 1997. On texture and its description of aerial image texture, Journal of Surveying
and Mapping, Vol.26,No.3, pp.222~227.
International Archives of Photogrammetry and Remote Sensing. Vol. XXXIII, Part B3. Amsterdam 2000. 1053