Related Experiment Video
Updated: Sep 25, 2025
06:41
Imaging and Analysis for Quantifying Maize (Zea mays) Abiotic Stress Phenotypes
Published on: March 28, 2025
1.1K
An Intelligent Analysis Method for 3D Wheat Grain and Ventral Sulcus Traits Based on Structured Light Imaging
Chenglong Huang1, Zhijie Qin1, Xiangdong Hua1
1College of Engineering, Huazhong Agricultural University, Wuhan, China.
Frontiers in Plant Science
|May 2, 2022
Summary
This study introduces a structured light imaging method for efficient 3D wheat grain phenotyping and ventral sulcus trait analysis. The developed technique accurately extracts key traits, aiding in wheat breeding research and yield prediction.
Area of Science:
- Agricultural Science
- Computer Vision
- Biotechnology
Background:
- Traditional manual measurement of wheat grain traits is inefficient, subjective, and labor-intensive.
- Ventral sulcus traits are crucial for wheat flour yield but require destructive measurement.
- Accurate 3D phenotypic data is vital for wheat yield and breeding programs.
Purpose of the Study:
- To develop an intelligent, non-destructive method for extracting 3D wheat grain phenotypes and ventral sulcus traits.
- To optimize experimental conditions for structured light imaging of wheat grains.
- To build and validate a wheat grain weight prediction model using phenotypic data.
Main Methods:
- Acquired 3D point cloud data using a structured light scanner.
- Developed algorithms for single grain segmentation and ventral sulcus localization.
- Conducted three-level orthogonal experiments to optimize scanning parameters (rotation angle, scanning angle, stage color).
- Extracted 28 3D phenotypic characters and 4 ventral sulcus traits.
- Constructed and validated a wheat grain weight model using 32 phenotypic traits.
Main Results:
- Optimized conditions for structured light imaging were identified: 30° rotation angle, 37° scanning angle, and black stage color.
- Mean absolute percentage errors (MAPEs) for grain length, width, thickness, and ventral sulcus depth were below 5%.
- The wheat grain weight prediction model achieved R-squared values ranging from 0.77 to 0.83.
- Developed specialized software integrating phenotype extraction and grain weight prediction.
Conclusions:
- The proposed structured light imaging method provides an efficient and effective approach for 3D wheat grain phenotyping and ventral sulcus trait analysis.
- This technology can significantly support wheat breeding research by enabling accurate, non-destructive measurements and predictive modeling.
- The integrated software facilitates practical application in agricultural research and development.
