地理空间信息2025，Vol.23Issue(4)：59-62,4.DOI:10.3969/j.issn.1672-4623.2025.04.012
植被指数引导的高光谱影像小麦种植区提取方法研究
Research on Wheat Planting Areas Extraction Method from Hyperspectral Images Guided by Vegetation Index
摘要
Abstract
Crops exhibit significant spectral characteristics in different bands,providing a natural advantage for crop classification in hyperspec-tral images.To address spatial information gaps and inconsistent spectral responses in crop extraction from hyperspectral remote sensing images,we proposed a crop extraction network that coupled multiple vegetation indices.We utilized vegetation indices to construct a spectral feature li-brary,assisting in crop extraction.By introducing a scale-aware attention mechanism and a cross-channel enhancement module,we addressed de-ficiencies in spatial description and channel correlation information.Taking wheat,a typical crop in Inner Mongolia,as the experimental subject,the result demonstrates the significant performance of this method in crop extraction.关键词
植被指数/高光谱影像/农作物提取/深度学习Key words
vegetation index/hyperspectral image/crop extraction/deep learning分类
天文与地球科学引用本文复制引用
田婷婷,杜斌,谭美淋,岳云峰,黄仡凡,张帅..植被指数引导的高光谱影像小麦种植区提取方法研究[J].地理空间信息,2025,23(4):59-62,4.基金项目
内蒙古自治区测绘地理信息中心自选项目(NMGZCS-C-F-220342). （NMGZCS-C-F-220342）
