Enhanced radiative transfer model boosts deep learning in plant phenotyping #PlantPhenotypeModeling

A research team has developed a radiative transfer modeling framework using Helios 3D plant modeling software to simulate various types of plant images with reference labels, reducing the need for manual data annotation. This framework enables efficient training of deep learning models for plant phenotyping, enhancing crop trait analysis and agricultural research. The study, published in Plant Phenomics, validated the model’s accuracy in simulating radiation absorption and reflection, with high precision scores and effective camera calibration results. The synthetic images generated by the model closely resemble real images, confirming its ability to produce high-quality annotated plant images. Lead researcher Tong Lei highlights the model’s capability to simulate complex plant and soil scenes, linking biophysical processes to plant properties. In summary, the study introduces a radiative transfer modeling framework for simulating plant images and improving deep learning model training for plant phenotyping. Future developments aim to enhance model flexibility and incorporate more complex processes for efficient analysis of plant traits and physiological states in high-throughput phenotyping and agricultural research.

Source link

Source link: https://phys.org/news/2024-07-framework-deep-phenotyping.html