We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
结合激光雷达和三维性状分析的田间小麦产量分级研究.
- Authors
施凌天; 丁国辉; 夏云鹏; 葛玉峰; 周济
- Abstract
[Objectives]Classifying yield potential in the field is an important research area in wheat breeding, cultivation and production. The objective of this study was to use LiDAR to conduct field phenotyping and dynamically monitoring of wheat canopy-level traits in order to provide phenotypic evidence to enable in-field yield classification. [Methods]We employed a backpack LiDAR system to collect 3D point clouds of wheat plants, based on which a 3D trait analysis pipeline could be established, including the correction of field-level 3D points, the plot-level point cloud segmentation, and the extraction of canopy-level traits such as crop height, canopy cover, 3D canopy surface area, and 3D canopy index (3DCI),from 486 wheat plots at key growth stages. [Results]The reliability of the analysis pipeline was verified using linear regression analysis with manual scoring, resulting in the coefficient of determination (R²) for all the traits (P<0.001,n=486 plots),including crop height (R²=0.866 0,RMSE=5.66 cm),canopy cover (R²=0.899 3,RMSE=0.057 4),3D canopy surface area (R²=0.836 4,RMSE=0.170 3),and 3DCI (R²=0.769 5,RMSE=0.265 5) . Then, we determined that early grain filling stage as the key phase for conducting yield classification due to the correlation analysis between post-harvest yield and key canopy-level traits, based on which yield classification, yield-based trait clustering, as well as related variety grouping were accomplished. [Conclusions]The algorithms presented in this study are capable of effectively extracting plot-level 3D canopy traits to classify yield production and phenotypes for high-yield wheat varieties, providing a reliable analytic approach to enable yield-related classification in breeding, cultivation, and agricultural production.
- Subjects
CLUSTER analysis (Statistics); LIDAR; WHEAT; BACKPACKS; CLASSIFICATION
- Publication
Journal of Nanjing Agricultural University / Nanjuing Nongye Daxue Xuebao, 2023, Vol 46, Issue 6, p1011
- ISSN
1000-2030
- Publication type
Article
- DOI
10.7685/jnau.202302018