提取天气类别动态特征的算法研究

Development of a Weather Recognition Algorithm Based on Dynamic Characteristics

现有的用于天气类别识别的动态特征存在一定不足,本文进行了相应改进,提高了整体的识别率。现有的相关性特征提取依赖于人工选取天气现象明显的区域进行,基于本文的改进算法能实现天气现象的自动识别。

Existingdynamic characteristics used to weather recognition is lack of accuracy,this study have made some modifications to increase the overall recognition rate.Existing correlation feature extraction depends on artificial select a region where weather phenomenon is obviously,however based on an improved algorithm in this paper can realized identify weather events automatically.