摘要
提出一种基于图割模型的卫星云图云检测方法。利用FY-2C卫星云图的长波红外通道和可见光通道的云图提取了10个灰度特征和80个Gabor纹理特征,再用主成分分析方法(principal component analysis,PCA)降维到9个主成分。将这9个主成分构成的特征作为每个像素的特征,建立相似度矩阵,再利用改进的Normalized Cuts模型进行分割,将云图分成了晴空区域和有云区域。与地面观测结果相比,平均一致率达到86.51%,表明将Gabor纹理特征和灰度特征相结合并利用改进的Normalized Cuts模型对卫星云图云检测有比较好的效果。
A novel approach is proposed for cloud detection in satellite cloud image based on the im- proved Graph Cuts Model. 10 gray features and 80 Gabor texture features were selected from two chan-nels of FY-2C satellite image（ infrared channel 1 and visible light channel）, and the dimensions of fea-ture vector were reduced to 9 by using PCA （principal component analysis）. Then the similarity matrix was built up by those feature vectors. By using the spectral graph theoretic framework of Normalized Cuts, the cloud image was divided into 2 parts, the clear sky and the cloudy sky. Compared with the re-sult of surface observation, the average consistency was 86.51%. The results demonstrate that this method is effective in cloud detection.
出处
《大气科学学报》
CSCD
北大核心
2012年第4期502-507,共6页
Transactions of Atmospheric Sciences
基金
国家自然科学基金资助项目(60802039
61071146)
