HJ-1 A高光谱数据的条带噪声去除方法研究  

A Strip Noise Removal Method of Hyperspectral Data HJ-1 A

作　　者：郭兴杰 王阳春 汪爱华 马卫胜 

机构地区：[1]北京宇视蓝图信息技术有限公司,北京100096 [2]中国兵器工业第二零八研究所,北京102202 [3]二十一世纪空间技术应用股份有限公司,北京100096

出　　处：《遥感信息》2011年第1期54-58,共5页Remote Sensing Information

基　　金：遥感信息柔性生产和快速服务平台建设(编号:D08080103490801)

摘　　要：针对环境减灾小卫星的高光谱图像条带的特点,提出了基于光谱空间连续性的倾斜条带去除方法。高光谱数据的光谱分辨率达到纳米级,光谱波段多,在一定范围内可以连续成像,具有光谱空间的连续性,基于光谱空间连续性的条带去除方法利用了光谱空间连续性的这一重要特点。本文在考虑了图像条带噪声的倾斜角度的基础上,成功地将该方法应用于批量的环境减灾小卫星2级高光谱数据,进行了相对辐射校正的研究,并将相邻列均衡方法应用于单幅环境减灾小卫星2级高光谱数据,对比二者的单幅图像条带去除效果,结果证明基于光谱空间连续性的条带去除方法较相邻列均衡方法更适合于对环境减灾小卫星的2级高光谱数据进行条带噪声的去除。Concerning the characteristics of strip in high-spectral image of HJ-1 satellite,the Inclined Strip Removal Method based on the continuity of spectral space is proposed in this article.Hyperspectral data has nanoscale spectral resolution,large number of spectral bands,and can be continuous imaged within a certain spectral range.These important features are used by the Inclined Strip Removal Method based on the continuity of spectral space.Taking into account the angle of the image strip noise,this method is applied to level 2 hyperspectral data of HJ-1 satellite successfully,and the relative radiometric correction is studied.Another method is applied to the data,too.Then the results are compared,and can be seen that the Inclined Strip Removal Method based on the continuity of spectral space is more suitable for the level 2 hyperspectral data of HJ-1 satellite.

关 键 词：高光谱 条带去除 光谱空间 环境减灾小卫星 

分 类 号：TP79[自动化与计算机技术—检测技术与自动化装置]