深度学习作物分类模型空间泛化能力分析  

Spatial generalization ability analysis of deep learning cropclassification models

作　　者：盖爽 张锦水 朱爽[5] GE Shuang;ZHANG JinShui;ZHU Shuang(State Key Laboratory of Remote Sensing Science,Beijing Normal University,Beijing 100875,China;Jiangsu Provincial Key Laboratory of Geographic Information Science and Technology,International Institute for Earth System Science,Nanjing University,Nanjing 210023,China;Beijing Engineering Research Center for Global Land Remote Sensing Products,Faculty of Geographical Science,Beijing Normal University,Beijing 100875,China;Institute of Remote Sensing Science and Engineering,Faculty of Geographical Science,Beijing Normal University,Beijing 100875,China;Beijing Polytechnic College,Beijing 100042,China)

机构地区：[1]北京师范大学地理科学学部遥感科学国家重点实验室,北京100875 [2]南京大学江苏省地理信息技术重点实验室/国际地球系统科学研究所,南京210023 [3]北京师范大学地理科学学部北京市陆表遥感数据产品工程技术研究中心,北京100875 [4]北京师范大学地理科学学部遥感科学与工程研究院,北京100875 [5]北京工业职业技术学院,北京100042

出　　处：《遥感学报》2023年第12期2796-2814,共19页NATIONAL REMOTE SENSING BULLETIN

基　　金：“高分辨率对地观测系统重大专项支持项目”民用部分(编号:20-Y30F10-9001-20/22)。

摘　　要：大数据驱动训练的深度学习模型是当今农作物分类的最新方法。当前研究仍然主要关注该模型方法的创新性,其在特定时间、特定地区的作物分类模型的泛化能力分析经常被忽略。因此,提高遥感分类模型在大尺度空间范围的有效迁移能力是遥感技术支撑地球系统科学研究和社会应用的关键。本研究通过设计实验分析了模型架构、作物物候特征、农区地块尺度、数据类型等因素对作物分类模型泛化能力的影响。结果表明:一方面当训练区和测试区地块大小发生明显变化时,MultiResUNet相对于SegNet,DeepLab V3+和U-Net具有更好的泛化性能。然而,单纯依靠MultiResUNet的泛化能力依然无法完全克服地块空间形态的变化对模型迁移的不利影响,为获得更高精度的华北玉米分布信息,需要优先使用与华北地区农业景观更相似的东北作物分布数据产品进行深度学习模型训练;另一方面,相对于TOA (Top of Atmosphere)数据,采用SR (Surface Reflectance)数据更有利于模型在跨洲际尺度进行空间迁移,因此,在大尺度作物制图研究中,应优先考虑使用SR数据。综上,本研究从一定程度上验证了影响农作物分类模型迁移性能的内在因素,可为大尺度作物制图提供科学依据。Timely and accurate global crop mapping is important for global food security assessment.However,existing crop classification models are often targeted at specific regions,and their performance in other regions has not been fully evaluated.This study determined the critical period of crop growth in different regions to realize the effective transfer of the model in large-scale regions,and the remote sensing data during these critical periods were filtered such that the same crop in different regions showed similar characteristics on these remote sensing images.This way helps the model achieve a better transfer effect.In this study,the MultiResUNet,SegNet,DeepLab V3+,and U-Net models were trained using data from Northeast China,and the optimal F1 value for summer corn recognition in the study areas in North China can reach more than 0.97.This research also analyzed the factors that affected the generalization ability of the model.The issues addressed in this article include(1)using existing crop distribution data products as the ground truth samples for model training to solve the problem of lack of training samples for the deep learning model.We compared and analyzed the applicability of the models trained using the US Cropland Data Layer and Northeast crop distribution data products in North China.(2)We compared the generalization performance of depth models with different architectures.(3)We compared and analyzed the influence of different data types on the generalization ability of the model.(4)We comparatively analyzed the impact of crop phenology changes on the generalization ability of the model.Results show that MultiResUNet has better generalization performance than other networks when the plot size in the training and test areas varies significantly However,the generalization ability of MultiResUNet alone still cannot completely overcome the adverse effect of the change in plot spatial morphology on model migration.The crop distribution data products in Northeast China,which are more similar to the agric

关 键 词：模型泛化 深度学习 SegNet DeepLab V3+ U-Net MultiResUNet 作物制图 

分 类 号：TP701[自动化与计算机技术—检测技术与自动化装置] P2[自动化与计算机技术—控制科学与工程]