基于BP神经网络的风云四号遥感图像云检测算法
投稿时间:2017-10-18  修订日期:2017-11-14  点此下载全文
引用本文:
摘要点击次数: 91
全文下载次数: 
作者单位E-mail
高军 上海海事大学 jungao@shmtu.ac.cn 
王恺 上海海事大学  
田晓宇 上海海事大学  
陈建 上海海事大学  
中文摘要:云检测是卫星遥感的一个重要应用,在于天气预报等多个方面发挥了关键的作用。多通道多阈值的方法常被应用于云检测的研究中,但是由于复杂多变的光谱特征以及难以有效表达的空间形态特征,导致阈值的大幅波动,且阈值的选取通常带有个人主观性,使得多通道多阈值云检测方法对于大范围复杂情况下的云难以取得稳定有效的检测效果。2016年12月,我国成功发射新一代静止气象卫星风云四号A星,该卫星拥有更多的通道以及更高的分辨率。如何充分利用风云四号提供的先进特性并且提高云检测效果是一个新的挑战。本文通过对风云四号相隔15分钟的图像进行分析,提出了归一化动云指数,加强数据集对低云、薄云及云系运动边缘的检测,并提出了一种基于动云归一化指数的动态阈值法用于初步云检测,然后通过初步结果进一步提出基于BP神经网络的云检测算法。实验结果表明,该算法可以消除阈值选取中的主观影响,在全天候大范围复杂下垫面的遥感图像数据中可以取得较好的云检测效果。
中文关键词:云检测,遥感图像处理,风云四号,BP神经网络
 
A BP-NN based Cloud Detection Method for FY-4 Remote Sensing Images
Abstract:Cloud detection is an important application of satellite remote sensing, which plays an essential role in weather forecast and other aspects. Multi-channel and multi-threshold method is often used in cloud detection research. However, the threshold is often fluctuated and is decided subjectively, so it is hard to achieve stable and effective detection result in large scale situations, due to the complex spectral characteristics and the spatial features that are difficult to be effectively expressed. Fengyun 4A (FY-4A), the Chinese new generation geostationary meteorological satellite was successfully launched in December 2016. FY-4A has more spectral channels and higher spatial resolution than the previous ones. It is a big challenge to make full use of the advanced characteristics of FY-4A in improving the results of cloud detection. In this paper, Normalized Difference Cloud Moving Index (NDCMI) is put forward by analyzing the remote sensing data every 15 minutes. By applying NDCMI, the detection of low cloud, thin cloud and the edge of moving cloud can be enhanced. We proposed a dynamic threshold method with NDCMI for preliminary cloud detection, and then through the preliminary results, we proposed a novel cloud detection method based on back propagation neural network (BP-NN). The experimental results show that the proposed method can eliminate the subjectivity in threshold selection, and can achieve better cloud detection results for remote sensing image in complex situations.
keywords:Cloud  Detection, Remote  Image Processing, FY-4, Back  Propagation Neural  Network
  查看/发表评论  下载PDF阅读器

版权所有:《红外与毫米波学报》编辑部