双主干深度学习网络支撑下的ZY1E影像耕地智能提取

doi:10.13474/j.cnki.11-2246.2026.0820

摘要/Abstract

摘要： [目的] 耕地是保障国家粮食安全和生态环境稳定的重要基础性资源。随着高分辨率卫星数据的快速发展，利用深度学习技术实现耕地高精度自动识别成为农业遥感研究的重要方向。[方法] 本文以江苏省宿迁地区为研究区，对资源一号02D（ZY1E）卫星影像进行辐射定标、大气校正、融合与几何配准等预处理操作，并通过多类型数据增强策略构建高质量训练样本集。在此基础上，提出了一种融合EfficientNet-B3与ResNet-50双主干网络的DB-DeepLabV3+语义分割模型。该模型中，EfficientNet-B3主干网络用于高效提取多尺度特征，ResNet-50主干增强深层语义表达能力，两者在DeepLabV3+框架下实现结构互补与边界优化。[结果] 结果表明，DB-DeepLabV3+模型在耕地识别中表现出较高的精度与稳定性，其总体精度(OA)达到90.91%，Kappa系数为0.818，交并比(IoU)为0.832，均优于单主干网络的DeepLabV3+、EfficientNet-B3+DeepLabV3+及ResNet-50+DeepLabV3+模型。模型在不同数据划分比例下仍保持良好的性能一致性，验证了其较强的泛化能力。[结论] 本文研究对于双主干结构在高分辨率遥感影像的耕地提取具有显著优势，为国产ZY1E卫星在农业智能识别与国土监测中的应用提供了有效技术支撑。

关键词: 耕地识别, 高分辨率遥感, 深度学习, ResNet-50, EfficientNet, DeepLabV3+

Abstract: [Purposes] Cultivated land is a fundamental resource crucial for ensuring national food security and ecological stability.With the rapid development of high-resolution satellite data,utilizing deep learning technology for high-precision automated identification of cultivated land has become a significant direction in agricultural remote sensing research. [Methods] This study takes the Suqian area of Jiangsu province as the research area.ZY1E images underwent preprocessing operations including radiometric calibration,atmospheric correction,fusion,and geometric registration.A high-quality training sample set was constructed through a multi-type data augmentation strategy.On this basis,a DB-DeepLabV3+semantic segmentation model integrating the dual-backbone networks of EfficientNet-B3 and ResNet-50 is proposed.In this model,the EfficientNet-B3 backbone network is used for efficient multi-scale feature extraction,while the ResNet-50 backbone enhances deep semantic representation capability； the two achieve structural complementarity and boundary optimization within the DeepLabV3+framework. [Findings] Results indicate that the DB-DeepLabV3+model demonstrates high accuracy and stability in cultivated land identification,achieving an overall accuracy (OA)of 90.91%, a Kappa coefficient of 0.818,and an intersection over union (IoU)of 0.832,outperforming the single-backbone DeepLabV3+,EfficientNet-B3+DeepLabV3+,and ResNet-50+DeepLabV3+models.The model maintained good performance consistency under different data split ratios,verifying its strong generalization ability. [Conclusions] The research findings demonstrate that the dual-backbone structure holds significant advantages for cultivated land extraction from high-resolution remote sensing imagery and provides effective technical support for the application of the domestic ZY1E satellite in agricultural intelligent recognition and land monitoring．

Key words: cultivated land identification, high-resolution remote sensing, deep learning, ResNet-50, EfficientNet, DeepLabV3+

中图分类号:

P237

周洪, 张国和, 游思琪, 王冉, 丁鹏飞. 双主干深度学习网络支撑下的ZY1E影像耕地智能提取[J]. 测绘通报, 2026, 0(8): 137-144.

Zhou Hong, Zhang Guohe, You Siqi, Wang Ran, Ding Pengfei. Intelligent cultivated land extraction from ZY1E imagery via dual-backbone deep learning network[J]. Bulletin of Surveying and Mapping, 2026, 0(8): 137-144.

导出引用管理器 EndNote|Reference Manager|ProCite|BibTeX|RefWorks

链接本文: http://tb.chinasmp.com/CN/10.13474/j.cnki.11-2246.2026.0820

http://tb.chinasmp.com/CN/Y2026/V0/I8/137

参考文献

[1] Liu Jiyuan,Kuang Wenhui,Zhang Zengxiang,et al.Spatiotemporal characteristics,patterns,and causes of land-use changes in China since the late 1980s[J].Journal of Geographical Sciences,2014,24(2):195-210．
[2] Zhou Yang,Li Xunhuan,Liu Yansui.Cultivated land protection and rational use in China[J].Land Use Policy,2021,106:105454．
[3] 巫志雄,李乔宇,王宗良,等.遥感影像耕地提取的研究进展与展望[J].山东农业科学,2024,56(12):163-170．
[4] 陈晓璐,王彦芳,张红梅,等.基于ESTARFM NDVI的察汗淖尔流域灌溉耕地提取方法研究[J].中国生态农业学报(中英文),2021,29(6):1105-1116．
[5] 宋宪强,梁钊雄,周红艺,等.基于决策树与时序NDVI变化检测的耕地撂荒遥感监测:以四川省凉山州普格县为例[J].山地学报,2021,39(6):912-921．
[6] 赵士肄,闫金凤,杜佳雪.基于面向对象结合随机森林模型的Sentinel-2A影像耕地信息提取[J].河南理工大学学报(自然科学版),2023,42(2):55-61．
[7] 陈玲玲,施政,廖凯涛,等.基于卷积神经网络的高分遥感影像耕地提取研究[J].农业机械学报,2022,53(9):168-177．
[8] 程锐,魏妍冰,陆苗,等.基于集成深度学习模型的耕地地块提取[J].中国农业资源与区划,2022,43(7):273-281．
[9] 李倩楠,张杜娟,潘耀忠,等.MPSPNet和UNet网络下山东省高分辨耕地遥感提取[J].遥感学报,2023,27(2):471-491．
[10] Persello C,Tolpekin V A,Bergado J R,et al.Delineation of agricultural fields in smallholder farms from satellite images using fully convolutional networks and combinatorial grouping[J].Remote Sensing of Environment,2019,231:111253．
[11] 毛星,金晶,张欣,等.基于改进DeepLabV3+模型和迁移学习的高分遥感耕地信息提取方法[J].江苏农业学报,2023,39(7):1519-1529．
[12] 肖茂池,林娜,李双桃,等.融合注意力机制的DeepLabV3+耕地遥感提取研究[J].遥感技术与应用,2025,40(5):1323-1332．
[13] 刘圳,刘德儿,赵尘.融合高分辨率网络与双分支结构的耕地范围建筑物遥感监测[J].遥感学报,2025,29(9):2776-2787．
[14] Wu Shikun,Su Yingyue,Lu Xiaojun,et al.Extraction and mapping of cropland parcels in typical regions of Southern China using unmanned aerial vehicle multispectral images and deep learning[J].Drones,2023,7(5):285．
[15] Chen L C,Zhu Yukun,Papandreou G,et al.Encoder-decoder with atrous separable convolution for semantic image segmentation[C]//Proceedings of 2018 Computer Vision-ECCV.Cham:Springer,2018:833-851．
[16] 李英龙,邓毓弸,孔赟珑,等.基于全色-多光谱双流卷积网络的端到端地物分类方法[J].自然资源遥感,2025,37(5):152-161．
[17] Alhichri H,Alswayed A S,Bazi Y,et al.Classification of remote sensing images using EfficientNet-B3 CNN model with attention[J].IEEE Access,2021,9:14078-14094．
[18] 张雷,丁雪,王金亮,等.一种耦合ResNet与Transformer的高分遥感语义分割方法[J].激光与光电子学进展,2025,62(16)：380-391．
[19] Ambar M K,Öztoprak H,Yurtkan K.Optimizing ResNet-50 for multiclass classification:a multi-stage learning approach[J].IEEE Access,2025,13:142517-142534．
[20] Wilson D R,Martinez T R.The need for small learning rates on large problems[C]//Proceedings of 2002 IJCNN'01.International Joint Conference on Neural Networks.Washington：IEEE,2002:115-119．
[21] Zhang Dujuan,Zhu Xiufang,Pan Yaozhong,et al.Comparative analysis of deep learning and traditional methods for high-resolution cropland extraction with different training data characteristics[J].Land,2025,14(10):2038．
[22] 田富有,曹玉佩,赵航,等.结合空间注意力机制与多任务学习的耕地地块分割模型[J].遥感学报,2024,28(11):2850-2864．
[23] Sharma N,Gupta S,Ali Mohammed Al-Yarimi F,et al.DBA-DeepLab:dual-backbone attention-enhanced DeepLabV3+model for plant disease segmentation[J].Food Science &Nutrition,2025,13(7):e70668．
[24] 张新长,黄健锋,宁婷.高分辨率遥感影像耕地提取研究进展与展望[J].武汉大学学报(信息科学版),2023,48(10):1582-1590．