ChinaWheat10 - 2018-2024年中国10米空间分辨率冬小麦识别数据集

南京农业大学更新发布了中国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。研究团队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，生产了2018-2024年10米空间分辨率的中国冬小麦空间分布数据集（ChinaWheat10）。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。

Nanjing Agricultural University has updated and released a 10-meter spatial resolution winter wheat identification dataset for China. Timely and accurate spatial distribution information of winter wheat is critical for food security and crop production management. Due to the high cost and low efficiency of acquiring training data, large-scale, high-quality, high-spatial-resolution winter wheat classification products remain scarce. The research team proposed an automated training data generation (ATDG) method that fuses winter wheat phenology, spectral and polarization information, to generate high-quality winter wheat training samples for remote sensing mapping of winter wheat based on machine learning approaches. Furthermore, the classification model was pre-trained using the generated training data, and then combined with the model transfer (MT) method to achieve intra-growth-season mapping of winter wheat. By integrating ATDG and MT, as well as optical and radar imagery, the ChinaWheat10 dataset, a 10-meter spatial resolution spatial distribution dataset of winter wheat in China from 2018 to 2024, was produced. The ATDG and MT methods have great potential for large-scale, high-resolution and high-precision automated mapping of winter wheat.