无线电工程2025，Vol.55Issue(8)：1675-1682,8.DOI:10.3969/j.issn.1003-3106.2025.08.015
基于NDVI时序特征分解的地表覆盖分类方法
Land Cover Classification Based on Time Series Feature Decomposition of NDVI
摘要
Abstract
Vegetation indices are critical features for distinguishing different land cover types,but they are significantly influenced by factors such as climate,atmospheric bidirectional reflectance,and geographic conditions.Existing land cover classification methods based on vegetation indices primarily utilize time-series vegetation indices but fail to account for the differential characteristics of vegetation index variations across seasons and trends among different land cover types.To address this limitation,a land cover classification method is proposed based on Normalized Difference Vegetation Index(NDVI)time-series feature decomposition.This approach employs time-series decomposition to extract seasonal and trend features from vegetation index sequences of land cover objects.These features are then weighted and combined using the Dynamic Time Warping(DTW)algorithm and the Induced Ordered Weighted Harmonic Averaging Operator(IOWHA).Finally,a land cover classification model is constructed based on the nearest neighbor algorithm.Classification experiments using MOD13Q1 data from Guangling County demonstrate that the proposed method effectively extracts differential features of vegetation indices across seasonal and trend variations for different land cover types.The overall classification accuracy exceeds 88%,representing an improvement of nearly 8 percentage points compared to methods without time-series feature decomposition.This approach provides a methodological foundation for time-series land cover classification studies based on vegetation index characteristics.关键词
动态时间规整/诱导有序加权调和平均算子/归一化植被指数/时间序列分解/地表覆盖分类Key words
DTW/IOWHA/NDVI/time series decomposition/land cover classification分类
信息技术与安全科学引用本文复制引用
葛艳琴,常巧梅..基于NDVI时序特征分解的地表覆盖分类方法[J].无线电工程,2025,55(8):1675-1682,8.基金项目
山西省基础研究计划联合资助项目(交控)(202303011222002) （交控）
高分辨率卫星数据服务课题(HX-202451) Jointly Funded Project of Shanxi Provincal Basic Research Program(Traffic Control)(202303011222002) （HX-202451）
High-resolution Satellite Data Service Project(HX-202451) （HX-202451）
