本文已被：浏览 773次 下载 1227次
中文摘要: 利用数学形态学方法对台风云系进行区域分割，计算并统计出了台风云系纹理特征的概率密度，然后结合模糊判别技术，利用多层神经网络对GMS云图的台风云系进行图象识别，结果表明，该方案是成功的。
中文关键词: 台风云系，概率密度，模糊判别，模糊神经网络
Abstract:Using the method of maths morphology,the area-spliting of typhoon clouds is conducted,and the possibility density of grain features of typhoon clouds is statistically analysed.Combined with the method of fuzzily discriminant and multi-level neural network,the typhoon clouds' picture recognition of GMS satellite image is studied. The result suggests thatthe method in successful.
文章编号： 中图分类号： 文献标志码：
基金项目:中国气象局科教司青年基金资助项目
| Author Name | Affiliation |
引用文本：
于波,冯民学,陈必云,1996.模糊神经网络在台风云系图象识别中的应用[J].气象,22(1):22-25.
,1996.The Application of Fuzzily Neural Network to Picture Recognition of Typhoon Clouds[J].Meteor Mon,22(1):22-25.
于波,冯民学,陈必云,1996.模糊神经网络在台风云系图象识别中的应用[J].气象,22(1):22-25.
,1996.The Application of Fuzzily Neural Network to Picture Recognition of Typhoon Clouds[J].Meteor Mon,22(1):22-25.