A dual-path model merging CNN and RNN with attention mechanism for crop classification

被引:2
作者
Zhang, Fuyao [1 ,2 ,3 ,4 ]
Yin, Jielin [1 ,2 ]
Wu, Nan [1 ,2 ]
Hu, Xinyu [1 ,2 ]
Sun, Shikun [1 ,2 ]
Wang, Yubao [1 ,2 ]
机构
[1] Northwest A&F Univ, Minist Educ, Key Lab Agr Soil & Water Engn Arid & Semiarid Area, Yangling 712100, Shaanxi, Peoples R China
[2] Northwest A&F Univ, Inst Water Saving Agr Arid Reg China, Yangling 712100, Shaanxi, Peoples R China
[3] Chinese Acad Sci, Inst Geog Sci & Nat Resources Res, Key Lab Land Surface Pattern & Simulat, Beijing 100101, Peoples R China
[4] Univ Chinese Acad Sci, Coll Resources & Environm, Beijing 100049, Peoples R China
基金
中国国家自然科学基金;
关键词
Crop classification; Deep learning; Google Earth Engine; Attention mechanism; Time-series data; INDEX; EXTENT; WATER;
D O I
10.1016/j.eja.2024.127273
中图分类号
S3 [农学(农艺学)];
学科分类号
0901 ;
摘要
Rapid and accurate crop classification is essential for estimating crop information and improving cropland management. The application of deep learning models for crop classification using time-series data has become the most promising method. However, most approaches rely on single models for data processing result in lower classification accuracy and poor stability. Therefore, this study proposes a dual-path approach with attention mechanisms (DPACR) to promote the performance of this model architecture in crop classification using time series data. Specifically, the model comprises two branches, the Recurrent neural network (RNN) branch with bidirectional gated recurrent units (GRU) with a self-attention mechanism, and the convolutional neural network (CNN) branch based on SE-ResNet. Crop classification is accomplished by a main classifier, supported by auxiliary classifiers from the two branches. Using the Google Earth Engine and the Sentinel-2 satellite data, DPACR was tested in the Hetao irrigation district in Inner Mongolia, China. The comparison experiment demonstrated that the DPACR achieved the highest overall accuracy (OA = 0.959) and Kappa coefficient (Kappa = 0.941) compared to other five models (MLP, SE-ResNet, Bi-At-GRU, SVM, and RF). DPACR excelled in classifying six crops, maintaining high accuracy across multiple classes. Compared to attention mechanisms, auxiliary classifiers can significantly improve classification performance. This study highlights the effective combination of cloud computing and deep learning for large-scale crop classification, providing a practical method for agricultural monitoring and management.
引用
收藏
页数:15
相关论文
共 41 条
  • [1] A high-performance and in-season classification system of field-level crop types using time-series Landsat data and a machine learning approach
    Cai, Yaping
    Guan, Kaiyu
    Peng, Jian
    Wang, Shaowen
    Seifert, Christopher
    Wardlow, Brian
    Li, Zhan
    [J]. REMOTE SENSING OF ENVIRONMENT, 2018, 210 : 35 - 47
  • [2] A joint learning Im-BiLSTM model for incomplete time-series Sentinel-2A data imputation and crop classification
    Chen, Baili
    Zheng, Hongwei
    Wang, Lili
    Hellwich, Olaf
    Chen, Chunbo
    Yang, Liao
    Liu, Tie
    Luo, Geping
    Bao, Anming
    Chen, Xi
    [J]. INTERNATIONAL JOURNAL OF APPLIED EARTH OBSERVATION AND GEOINFORMATION, 2022, 108
  • [3] Annual maps of global artificial impervious area (GAIA) between 1985 and 2018
    Gong, Peng
    Li, Xuecao
    Wang, Jie
    Bai, Yuqi
    Cheng, Bin
    Hu, Tengyun
    Liu, Xiaoping
    Xu, Bing
    Yang, Jun
    Zhang, Wei
    Zhou, Yuyu
    [J]. REMOTE SENSING OF ENVIRONMENT, 2020, 236
  • [4] Google Earth Engine: Planetary-scale geospatial analysis for everyone
    Gorelick, Noel
    Hancher, Matt
    Dixon, Mike
    Ilyushchenko, Simon
    Thau, David
    Moore, Rebecca
    [J]. REMOTE SENSING OF ENVIRONMENT, 2017, 202 : 18 - 27
  • [5] Improved Gaussian mixture model to map the flooded crops of VV and VH polarization data
    Guan, Haixiang
    Huang, Jianxi
    Li, Li
    Li, Xuecao
    Miao, Shuangxi
    Su, Wei
    Ma, Yuyang
    Niu, Quandi
    Huang, Hai
    [J]. REMOTE SENSING OF ENVIRONMENT, 2023, 295
  • [6] Hansen MC, 2013, SCIENCE, V342, P850, DOI [10.1126/science.1244693, 10.1126/science.1215904]
  • [7] Deep Residual Learning for Image Recognition
    He, Kaiming
    Zhang, Xiangyu
    Ren, Shaoqing
    Sun, Jian
    [J]. 2016 IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR), 2016, : 770 - 778
  • [8] Evaluation and Comparison of Random Forest and A-LSTM Networks for Large-scale Winter Wheat Identification
    He, Tianle
    Xie, Chuanjie
    Liu, Qingsheng
    Guan, Shiying
    Liu, Gaohuan
    [J]. REMOTE SENSING, 2019, 11 (14)
  • [9] DualNet: Learn Complementary Features for Image Recognition
    Hou, Saihui
    Liu, Xu
    Wang, Zilei
    [J]. 2017 IEEE INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV), 2017, : 502 - 510
  • [10] Hu J, 2018, PROC CVPR IEEE, P7132, DOI [10.1109/CVPR.2018.00745, 10.1109/TPAMI.2019.2913372]