双光谱卫星云图的模糊推理云分类

Cloud Fuzzy Inference and Classification based on Double-Spectrum Satellite Images

基于观测事实和分类经验,在对卫星云图上云系形态特征定性分析的基础上,采用模糊推理方法,提取出云类识别的诊断判别因子和模糊推理规则。通过对判别因子的模糊化处理和隶属度函数的调制优化,建立了卫星云图定量化的模糊推理云分类模型。4种常见云系的模糊推理分类试验结果表明,模糊推理方法在定量性、客观性和自动化等方面优于传统卫星云图目视判读方法,在卫星遥感资料的分析处理和天气预报等领域有较好的实用意义。

Based on the observed facts and actual experience as well as the qualitative analysis and diag-noses on the cloud structure characters of satellite images,some judgement factors and fuzzy inference rulesfor classifying cloud were picked-up by using fuzzy inference method with fuzzy/defuzzy treatment on thejudgement factors and optimization adjustment on the membership functions,a quantitative fuzzy inferencemodel for classifying cloud of satellite images was established. The results of 4 kinds of common cloud'sfuzzy inference classification experiments show that the fuzzy inference method is superior to traditionaldiscriminance in the aspect of quantification, objectivity and automatization,which promises to be appliedin the fields of satellite remote sensing data treatment and weather forecast.