Title :
Extracting the rotation feature of the developing typhoon
Author :
Wang, Ping ; Yang, Pei-Long ; Li, Wen-Li ; Lu, Han-Qing
Author_Institution :
Inst. of Autom., Tianjin Univ., China
Abstract :
Based on infrared and visible time series satellite images, this paper studies the unique rules of the developing typhoon, and constructs the dynamic rotation feature and the rotation feature´s durative feature grounded on the cloud motion wind. Meanwhile, an effective algorithm has been proposed to extract the rotation feature. Experimental results show that the rotation feature with favorable running qualities could effectively recognize the developing typhoon.
Keywords :
disasters; image processing; infrared imaging; storms; time series; cloud motion wind; infrared image; typhoon development; visible time series satellite images; Developing typhoon; cloud motion wind vectors; persistent feature; rotation feature;
Conference_Titel :
Machine Learning and Cybernetics, 2005. Proceedings of 2005 International Conference on
Conference_Location :
Guangzhou, China
Print_ISBN :
0-7803-9091-1
DOI :
10.1109/ICMLC.2005.1527867