Abstract: Automatic cloud detection is the first step of using the remote sensing data. In this study, a novel cloud detection method based on Logistic regression model is proposed. The stepwise regression is applied to extract the feature from the satellite image of FY-2C. The Logistic regression model is applied to detect the cloud. The result of experiment, compared with the surface observations, shows that the Logistic regression model is effective for cloud image processing, and the method is more accurate than the traditional threshold algorithm.

Key words: Logistic regression, satellite image, cloud detection, feature extract

摘要： 云的自动检测和分类识别是所有卫星遥感资料应用的第一个步骤。基于Logistic回归模型的云图处理方法被用于FY-2C卫星云图的处理。利用逐步回归方法对云图的灰度及纹理特征进行提取，并计算出每个特征的回归系数；利用提取的特征进行云检测实验。将实验结果与地面观测资料进行对比，表明Logistic回归模型对云图处理是有效的，并且与传统的动态阈值分割方法相比，云检测的效果更好。

关键词: Logistic回归, 卫星云图, 云检测, 特征提取