SOFM神经网络的FY-3A/VIRR多光谱图像云相态反演方法  

A Cloud Phase Retrieval Approach Based on SOFM Neural Network Using FY-3A/VIRR Multi-channel Images

作　　者：郭晶[1,2] 杨春平[1] 叶玉堂[1] 饶长辉[2] 

机构地区：[1]电子科技大学光电信息学院,成都610054 [2]中国科学院光电技术研究所,成都610209

出　　处：《光电工程》2015年第12期20-24,共5页Opto-Electronic Engineering

基　　金：国家自然科学基金资助项目(11173008);中央高校基本科研业务费专项资金资助项目(103.1.2E022050205)

摘　　要：针对使用阈值方法反演云相态存在的不足,本文提出了一种基于Self-Organizing Feature Map(SOFM)神经网络的云相态反演方法。采用覆盖中国地域的Feng Yun-3A/Visible and Inf Rared Radiometer(FY-3A/VIRR)多光谱图像开展了云相态反演实验。实验结果表明:SOFM神经网络方法与K-means方法的结果具有较好的一致性,且SOFM神经网络方法反演云相态的准确性优于FY-3A业务产品。此外,SOFM神经网络方法反演云相态所需时间仅为FY-3A业务产品的约1/3。To address problems of cloud phase retrieval using the threshold method, a cloud phase retrieval approach based on Self-Organizing Feature Map（SOFM） neural network was proposed. Cloud phase retrieval experiments were conducted using Feng Yun-3A/Visible and Inf Rared Radiometer（FY-3A/VIRR） multi-channel images, which cover the China＇s territory. Experiment results indicated that the results from the SOFM neural network approach and the K-means method have good consistency, and the retrieval accuracy of the SOFM neural network exceeds that of the FY-3A operational product. Additionally, retrieval time consumed by the SOFM neural network approach is only about one third of that of the FY-3A operational product.

关 键 词：人工神经网络 FY-3A/VIRR 云相态 阈值方法 业务产品 

分 类 号：P407.4[天文地球—大气科学及气象学]