【摘要】 针对绿潮遥感信息提取过程中容易出现的几种易混淆因素,开展了多源卫星绿潮遥感信息提取易混淆因素分析研究。基于多源遥感卫星图像,分析了光学和微波遥感数据在提取绿潮过程中常见的几种易混淆因素。结果发现:(1)HJ-1卫星CCD遥感影像上,岛屿、船只、堤坝、云都是易混淆因素。在信息提取中,需结合基础地理资料或"天地图",将岛屿识别出来,此方法同样适用于MODIS和SAR数据。对于堤坝、船只和有云覆盖的绿潮区域,则需要通过人机交互的方式进行识别。(2)MODIS遥感影像中散布的小面积云和条带噪声是易混淆因素,因此需在MODIS数据预处理中进行云掩膜和条带噪声去除。(3)ENVISAT ASAR遥感影像中船只是易混淆因素,需通过人机交互的方式进行区分。更多还原

【Abstract】 Based on multi-source satellite images,the common confusable factors in retrieving green tide from optical and microwave remote sensing data were analyzed.Results show that:(1)With respect to CCD images of HJ-1satellite,islands,ships,dams and clouds are the confusable factors.So when we extract the green tide information from the remote sensing images,we must combine them with basic geographic data or "Map World",which can help us identify the islands from the images.This method also applies to MODIS and SAR data.For the dams,ships and the green tide area covered under the cloud,one needs to identify through the way of human-computer interaction.(2)For MODIS data,dispersed clouds with small area and stripe noise are confusable factors.When we extract the green tide information from MODIS remote sensing images,we need to mask the cloud and remove the stripe noise in preprocessing.(3)On ENVISAT ASAR data,ships are confusable factor.We can distinguish them by human-computer interaction.更多还原