WHU-Hi
收藏资源简介:
WHU-Hi是由武汉大学创建的高空间分辨率无人机载高光谱图像(H2)数据集,旨在解决现有数据集在空间分辨率、标记像素比例和子类别区分度方面的不足。该数据集包含WHU-Hi-LongKou、WHU-Hi-HanChuan和WHU-Hi-HongHu三个子集,分别在湖北省的不同农业区域采集,具有从分米级到厘米级的高空间分辨率,标记像素比例超过69%,包含多种农作物类型。数据集创建过程中,使用了Headwall Nano-Hyperspec传感器进行数据采集,并通过辐射校正和几何校正进行预处理。WHU-Hi数据集主要应用于精准农业和土地利用监测,以提高高光谱图像分类的准确性和效率。
WHU-Hi is a high spatial resolution unmanned aerial vehicle (UAV)-borne hyperspectral image (H2) dataset developed by Wuhan University, which aims to address the shortcomings of existing datasets in terms of spatial resolution, labeled pixel proportion, and subcategory distinguishability. This dataset comprises three subsets, namely WHU-Hi-LongKou, WHU-Hi-HanChuan, and WHU-Hi-HongHu, which were collected from different agricultural regions in Hubei Province. It features high spatial resolutions ranging from decimeter-level to centimeter-level, with a labeled pixel proportion exceeding 69%, and covers a variety of crop types. During the dataset construction, data was collected using the Headwall Nano-Hyperspec sensor, and preprocessed via radiometric correction and geometric correction. The WHU-Hi dataset is primarily applied in precision agriculture and land use monitoring to improve the accuracy and efficiency of hyperspectral image classification.




