2020年中国10米空间分辨率冬小麦识别数据集

及时准确的冬小麦空间分布信息对于粮食安全和作物生产管理至关重要。由于训练数据的获取成本高、效率低，大尺度、高质量、高空间分辨率的冬小麦分类产品依然匮乏。因此，我们提出了一种融合冬小麦物候、光谱和极化信息的训练数据自动化提取方法（Automated training data generation, ATDG），用于生成高质量的冬小麦训练样本，从而开展基于机器学习方法的冬小麦遥感制图。此外，基于生成的训练数据预训练分类模型，然后结合模型年际迁移方法（Model transfer, MT）实现了冬小麦生长季内制图。通过联合ATDG和MT，结合光学和雷达影像，我们生产了2020-2023年10米空间分辨率的中国冬小麦制图数据集（ChinaWheat10）。其中，2020、2021及2023年冬小麦分布图由ATDG方法生成，2022年分布图由MT方法生成。实地调查数据表明ChinaWheat10的总体精度在94%以上，在省市两级与统计数据的相关性（R2)分别在0.95和0.91以上，而且ChinaWheat10数据集中冬小麦田块细节明显。ATDG和MT方法在大尺度高分辨率高精度冬小麦自动化制图中具有巨大潜力。分类代码：1，冬小麦；0，非冬小麦。

Timely and accurate spatial distribution information of winter wheat is critical for food security and crop production management. Due to the high acquisition cost and low efficiency of training data, large-scale, high-quality, high-spatial-resolution winter wheat classification products remain scarce. Therefore, we propose an automated training data generation (ATDG) method that fuses phenological, spectral and polarization information of winter wheat to generate high-quality training samples for winter wheat, enabling remote sensing mapping of winter wheat using machine learning approaches. In addition, we pre-trained a classification model with the generated training data, and then combined it with the inter-annual model transfer (MT) method to achieve in-season winter wheat mapping. By integrating ATDG and MT along with optical and radar imagery, we produced a 10-meter spatial resolution winter wheat mapping dataset of China (ChinaWheat10) for the period 2020–2023. Specifically, the winter wheat distribution maps for 2020, 2021 and 2023 were generated via the ATDG method, while the map for 2022 was generated using the MT method. Field validation data demonstrate that ChinaWheat10 achieves an overall accuracy of over 94%. Its coefficients of determination (R²) with statistical data at the provincial and municipal levels are above 0.95 and 0.91, respectively, and the dataset features distinct details of winter wheat field parcels. The ATDG and MT methods hold great potential for large-scale, high-resolution, high-precision automated winter wheat mapping. Classification codes: 1 = winter wheat; 0 = non-winter wheat.

搜集汇总

背景与挑战

背景概述

该数据集是2020年中国10米空间分辨率冬小麦识别数据集，属于ChinaWheat10系列，采用自动化训练数据生成和模型迁移方法，结合光学和雷达遥感影像，实现了高精度冬小麦空间分布制图。数据集覆盖中国范围，总体精度超过94%，数据格式为栅格图像，适用于农学与遥感领域的作物识别研究。