多阈值及连通域距离权重的雾判识  

Fog Detection by Multi-threshold and Distance Weights of Connected Component

作　　者：李彬[1,2] 卢士庆 王飞[3] 孙小龙 张迎杰 LI Bin;LU Shiqing;WANG Fei;SUN Xiaolong;ZHANG Yingjie(Inner Mongolia Eco-and Agro-Meteorological Center,Hohhot 010051,China;Inner Mongolia Data and Application Center of High Resolution Earth Observation System,Hohhot 010051,China;Cloud-precipitation Physics and Weather Modification Key Laboratory,CMA,Beijing 100081,China;Chifeng Meteorological Bureau,Chifeng,Inner Mongolia 024000,China)

机构地区：[1]内蒙古自治区生态与农业气象中心,呼和浩特010051 [2]高分辨率对地观测系统内蒙古数据与应用中心,呼和浩特010051 [3]中国气象局云降水物理与人工影响天气重点开放实验室,北京100081 [4]赤峰市气象局,内蒙古赤峰024000

出　　处：《遥感信息》2022年第6期41-47,共7页Remote Sensing Information

基　　金：国家自然科学基金项目(41875181);风云卫星先行计划(FY-APP-2021.0405);内蒙古自治区科技计划(2020GG0016);内蒙古自治区自然科学基金项目(2022MS04015);内蒙古自治区气象局科技创新项目(nmqxkjcx202320)。

摘　　要：以往针对大雾遥感监测开展了许多研究,阈值法仍是主流研究方向。目前主要的难点是雾与低云的区分、轻雾及云下雾判识,或者多针对海雾,缺乏海陆通用的算法。文章利用Himawari-8卫星数据提出一种针对性识别方法,引入以多阈值判识为基础的连通域距离权重来判识云下雾。例举了6次大雾过程的判识结果,并结合卫星与地面观测数据进行了验证和精度分析。结果表明,该方法对海、陆大雾判识有较高精度,并在有限引入误差的情况下,可以实施一些云下雾的判识。文章为静止气象卫星雾遥感监测业务提供了有利参考。Many researches on fog detection of remote sensing have been carried out and have been succeed. The method by thresholds is still the main researching direction. At present, the main difficulties are the distinction between fog and stratus, and the identification of mist and fog under cloud. Most methods are for sea fog, and it is lack of common algorithms for sea and land. In this paper, a new method is proposed based on the data of Himawari-8 satellite, which is based on multi-thresholds and connected component of distance weight to identify the fog beneath the cloud. In this paper, the detected results of six samples are given, and the validation and accuracy analysis are carried out by combining interpretation and ground observation. The results show that the method has high accuracy in pure fog identification, and can effectively detect fog under cloud with limited errors. It provides a new idea for fog extraction and a favorable way for fog detection of geostationary meteorological satellite by real-time automation.

关 键 词：大雾检测指数 连通域 灰度熵 Himawari-8 距离权重 

分 类 号：TP79[自动化与计算机技术—检测技术与自动化装置]