结合MoCo自监督学习的多时相遥感农作物提取方法

By using MoCo self-supervised learning and the Swin-Unet model, the problems of high data annotation cost and insufficient utilization of temporal information in multi-temporal deep learning crop classification models are solved. This achieves high-precision crop classification and reduces reliance on manual annotation, making it suitable for refined management of crops using multi-temporal remote sensing.

CN121545065BActive Publication Date: 2026-07-17JILIN AGRICULTURAL UNIV
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
JILIN AGRICULTURAL UNIV
Filing Date
2026-01-21
Publication Date
2026-07-17

Smart Images

  • Figure CN121545065B_ABST
    Figure CN121545065B_ABST
Patent Text Reader

Abstract

结合MoCo自监督学习的多时相遥感农作物提取方法。涉及农业资源监测技术领域，具体涉及结合MoCo自监督学习的多时相遥感农作物提取方法。利用无标签卫星影像，采用MoCo自监督学习对空间编码器进行无标签预训练，使作物分类模型在有限的标签数据下仍能够学习到具有判别力的时空特征表征，从而有效缓解标注样本不足带来的分类精度问题。所述方法包括如下步骤：获取多时相农作物无标签遥感卫星影像数据集和有标签数据集；构建时相作物分类模型：MoCo自监督学习预训练空间特征编码器；采用有标签数据集，对作物分类模型进行有监督学习，得到最终时相作物分类模型；通过最终时相作物分类模型，对多时相农作物进行分类。
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Global land evapotranspiration space-time fusion method based on deep learning

    CN117010262A

  • Sparse sample crop drawing method based on weak supervision semantic segmentation

    CN119205976A