Soybean yield estimation and lodging discrimination based on lightweight UAV and point cloud deep learning
文献类型： 外文期刊
作者： Zhou, Longyu 1 ; Han, Dezhi 2 ; Sun, Guangyao 3 ; Liu, Yaling 4 ; Yan, Xiaofei 2 ; Jia, Hongchang 2 ; Yan, Long 5 ; Feng, Puyu 1 ; Li, Yinghui 6 ; Qiu, Lijuan 6 ; Ma, Yuntao 1 ;
作者机构： 1.China Agr Univ, Coll Land Sci & Technol, Beijing 100193, Peoples R China
2.Heilongjiang Acad Agr Sci, Heihe Branch, Heihe, Peoples R China
3.China Agr Univ, Coll Informat & Elect Engn, Beijing 100193, Peoples R China
4.Inner Mongolia Pratacultural Technol Innovat Ctr C, Hohhot, Inner Mongolia, Peoples R China
5.Inst Cereal & Oil Crops, Hebei Acad Agr & Forestry Sci, Shijiazhuang 050035, Hebei, Peoples R China
6.Chinese Acad Agr Sci, Inst Crop Sci, State Key Lab Crop Gene Resources & Breeding, Beijing 100081, Peoples R China
关键词： Remote sensing; 3D reconstruction; Point cloud; Multi-task learning; Digital image
期刊名称：PLANT PHENOMICS （ 影响因子：6.4； 五年影响因子：7.1 ）
ISSN： 2643-6515
年卷期： 2025 年 7 卷 2 期
页码：
收录情况： SCI
摘要： The unmanned aerial vehicle (UAV) platform has emerged as a powerful tool in soybean (Glycine max (L.) Merr.) breeding phenotype research due to its high throughput and adaptability. However, previous studies have predominantly relied on statistical features like vegetation indices and textures, overlooking the crucial structural information embedded in the data. Feature fusion has often been confined to a one-dimensional exponential form, which can decouple spatial and spectral information and neglect their interactions at the data level. In this study, we leverage our team's cross-circling oblique (CCO) route photography and Structure-from-Motion with Multi-View Stereo (SfM-MVS) techniques to reconstruct the three-dimensional (3D) structure of soybean canopies. Newly point cloud deep learning models SoyNet and SoyNet-Res were further created with two novel data-level fusion that integrate spatial structure and color information. Our results reveal that incorporating RGB color and vegetation index (VI) spectral information with spatial structure information, leads to a significant reduction in root mean square error (RMSE) for yield estimation (22.55 kg ha-1) and an improvement in F1-score for five-class lodging discrimination (0.06) at S7 growth stage. The SoyNet-Res model employing multi-task learning exhibits better accuracy in both yield estimation (RMSE: 349.45 kg ha-1) when compared to the H2O-AutoML. Furthermore, our findings indicate that multi-task deep learning outperforms single-task learning in lodging discrimination, achieving an accuracy top-2 of 0.87 and accuracy top-3 of 0.97 for five-class. In conclusion, the point cloud deep learning method exhibits tremendous potential in learning multi-phenotype tasks, laying the foundation for optimizing soybean breeding programs.
- 相关文献
作者其他论文 更多>>
-
A comparison of Chinese soybean collection with European soybean collections on genetic diversity by genome-wide scan
作者：Xu, Jiangyuan;Lu, Yuqing;Gu, Yongzhe;Gao, Huawei;Liu, Zhangxiong;Qiu, Lijuan;Yao, Xindong;Yao, Xindong;Leopold, Rittler;Yuan, Ming;Zhang, Yong;Sun, Rujian;Xue, Yongguo;Liu, Yeli;Han, Dezhi;Wang, Jinxing;Qiu, Lijuan
关键词：Chinese soybean; European soybean; Genetic diversity; Selective sweep analysis; Selected region
-
Genome-Wide Identification of GmPIF Family and Regulatory Pathway Analysis of GmPIF3g in Different Temperature Environments
作者：Liang, Xuefeng;Zhao, Caitong;Cui, Jiayang;Chen, Qingshan;Yang, Mingliang;Jiang, Zhenfeng;Liu, Zhihua;Han, Dezhi
关键词：soybean; PIF; bioinformatics analysis; biofilm interference; interacting protein
-
Soybean yield estimation and lodging classification based on UAV multi-source data and self-supervised contrastive learning
作者：Zhou, Longyu;Chen, Haochong;Feng, Puyu;Ma, Yuntao;Zhang, Yong;Wang, Lei;Li, Mingxue;Sun, Xuhong;Sun, Guangyao;Yan, Long;Qiu, Lijuan;Li, Yinghui
关键词：Remote sensing; 3D reconstruction; Point cloud; Multispectral; Structure from motion
-
Bridging the gap between hyperspectral imaging and crop breeding: soybean yield prediction and lodging classification with prototype contrastive learning
作者：Sun, Guangyao;Han, Shiteng;Xu, Yun;Ma, Yuntao;Zhang, Yong;Wang, Lei;Zhou, Longyu;Fei, Shuaipeng;Xiao, Shunfu;Ma, Yuntao;Yan, Long;Che, Yingpu;Li, Yinghui;Qiu, Lijuan
关键词：Unmanned Aerial Vehicle; Deep Learning; Resnet; Soybean Yield and Lodging Prediction
-
Identification of Quantitative Trait Loci for Node Number, Pod Number, and Seed Number in Soybean
作者：Zhang, Chunlei;Zha, Bire;Yuan, Rongqiang;Zhao, Kezhen;Liu, Xiulin;Wang, Xueyang;Zhang, Fengyi;Ren, Honglei;Zha, Bire;Sun, Jianqiang;Sun, Jianqiang;Qiu, Lijuan;Zhang, Bixian;Lamlom, Sobhi F.;Lamlom, Sobhi F.
关键词：soybean; differential expression analysis; SLAF sequencing; QTL mapping
-
Soybean Lodging Classification and Yield Prediction Using Multimodal UAV Data Fusion and Deep Learning
作者：Xu, Xingmei;Fang, Yushi;Yu, Helong;Ma, Yuntao;Sun, Guangyao;Guo, Yan;Ma, Yuntao;Zhang, Yong;Wang, Lei;Chen, Chen;Ren, Lisuo;Meng, Lei;Li, Yinghui;Qiu, Lijuan
关键词：UAV; deep learning; lodging classification; yield; data fusion
-
Developing a highly efficient transgenic approach to generate sheepgrass germplasm with enhanced biomass yield and salt tolerance
作者：Zhao, Haixia;Jiang, Xiping;Yang, Ruijuan;Wu, Zhenying;Fu, Chunxiang;Zhao, Haixia;Jiang, Xiping;Yang, Ruijuan;Wu, Zhenying;Fu, Chunxiang;Zhao, Haixia;Wu, Zhenying;Fu, Chunxiang;You, Jia;Wang, Jianli;Shen, Zhongbao;Liu, Yaling;Yuan, Feng;Wang, Yi;Zheng, Wendong
关键词：
