一种NDVI时间序列离散Fourier重建方法  

A Reconstruction Method of Time Series NDVI Based on Discrete Fourier Analyses

作　　者：马超[1] 郭增长[1] 张晓克[2] 

机构地区：[1]河南理工大学矿山空间信息技术国家测绘局重点实验室,河南焦作454000 [2]太原理工大学矿业工程学院,山西太原030024

出　　处：《测绘科学技术学报》2011年第5期347-350,355,共5页Journal of Geomatics Science and Technology

基　　金：国家973重点基础研究发展资助项目(2009CB226107);国家杰出青年基金资助项目(50525414);山西省科技攻关资助项目(20080311085);河南理工大学博士基金资助项目(B2009-46)

摘　　要：时间序列重建是多时相遥感变化监测研究的基础。多时相遥感数据由于气象、轨道周期等原因,很难达到严格意义的同时相。不同时相的特征信息不具有可比性。以一矿区归一化差值植被指数为例,采用2004-2007年SPOT II/IV数据,通过大气校正,波段运算,获得7个时相的包括伪不变特征区和矿业扰动区8个实验工作面的NDVI最大值、均值和振幅;进而采用快速Fourier变换的Fourier 4模型将NDVI转化为振幅,将物候信息转化为相位。实现了低数据样本下高时间、高空间分辨率NDVI时间序列数据集重建。该方法可为多时相遥感信息时间序列归一化、多源数据时间序列同化以及特征信息协同反演提供一些借鉴。Reconstructed NDVI time series is a research base of multi-temporal remote sensing change detection.Multi-temporal remote sensing data due to climate,orbital period and other reasons,it is difficult to achieve in the strict sense of the same time,and the characteristics information of different time is not comparable.A cased NDVI of coal mining regions,the maximum NDVI,mean NDVI and NDVI amplitude before and after the mining subsidence in Pseudo-Invariant Features Region（PIFs） and eight experimental faces of Luan mining subsidence area were extracted from 2004-2007 seven temporal remote sensing images（SPOT II/IV satellites） after atmospheric correction operation and band calculation.Then by means of the Fourier 4 model of Fast Fourier Transform（FFT）,the NDVI value was transformed into amplitude and the phonological was changed into phase information.A NDVI time series of high spatial-temporal resolution was reconstructed in the case of low data samples.The method maybe provides some reference to reconstruct time series of multi-temporal remote sensing information,to assimilate time series of multi-source data,and cooperative to inverse multi-characteristic information.

关 键 词：开采沉陷区 时间序列重建 离散Fourier分析 快速FOURIER变换 归一化差值植被指数 

分 类 号：P237[天文地球—摄影测量与遥感] TP028.8[天文地球—测绘科学与技术]