基于隐马尔可夫模型的强对流天气识别

Recognition of Severe Convective Weather Based on Hidden Markov Models

多普勒天气雷达是进行强对流天气识别的主要手段。应用图像处理技术,设计实现了多普勒雷达反射率图中单体的分割和特征提取算法。为有效利用单体特征的时间序列特性,提出了一种基于隐马尔可夫模型的强对流天气识别方法。实验表明,相比PUP系统,该方法的识别率更高,误报率更低。

Doppler weather radar is the primary means for recognizing the severe convective weather. With the application of image processing technology, a cells＇ segmentation and feature extraction algorithm from the Doppler radar reflectivity image was designed and implemented. For using the time series properties of cells＇ features effectively, a recognition method of the severe convective weather based on hidden Markov models was presented. Compared to PUP, the recognition rate of this method is higher and the false alarm rate is lower.