基于多纹理特征和PCA的地基云图识别研究

地基云图自动分类识别对于天气现象的诊断和预报具有重要意义。以地基云图为研究目标，首先在提取云图灰度共生矩阵和Gabor变换特征的基础上计算云图的多纹理组合特征，然后采用主成分分析法对组合特征进行数据降维，获取最终鉴别特征。通过对积雨云，高积云和层积云三类地基云图进行分类识别的实验结果表明，该方法可以同时提高云图分类的精度和效率。

Research on recognition of ground-based cloud images based on multi-texture features and PCA

Ground-based Cloud images automatic classification and identification plays an essential role in diagnosis and prediction of the weather phenomenon and has received great concern in recent years. In this paper, two different texture feature sets were extracted and combined from the ground-based Cloud images using the algorithms of Gray level co-occurrence matrix and Gabor transform. Then the principal component analysis (PCA) was used to reduce the data dimension of the combination of texture features, to get the final identification feature set. The recognition experimental results of cumulonimbus, altocumulus and stratocumulus show that the proposed method can improve the accuracy and efficiency of the classification of ground-based cloud images simultaneously.