Bio-Master: Design and Validation of a High-Throughput Biochemical Profiling Platform for Crop Canopies

Area of Science:

• Agricultural Science
• Plant Physiology
• Remote Sensing Technology

Background:

• Accurate crop biochemical profiling is essential for diagnosing plant physiological status.
• Conventional destructive methods are reliable but labor-intensive.
• Nondestructive techniques offer efficiency but often lack precision due to environmental and structural complexities.

Purpose of the Study:

• To develop and validate the Bio-Master phenotyping system for simultaneous, accurate, and efficient measurement of key crop biochemical components.
• To establish a robust estimation model for biochemical content using hyperspectral data.
• To compare the system's performance against conventional and existing nondestructive methods.

Main Methods:

• The Bio-Master system segments plant samples, measures hyperspectral reflectance and fresh weight in a controlled environment.
• An embedded Gaussian process regression model estimates dry matter, water, chlorophyll, and nitrogen content.
• A comprehensive dataset of rice varieties, nitrogen levels, and growth stages was used for model training and validation.

Main Results:

• The Bio-Master system accurately estimates biochemical components at both leaf and plant scales using leave-one-out validation.
• Each rice plant measurement takes approximately 5 minutes, providing multiple data points across the vertical profile.
• The system enables immediate, in-situ measurements, preserving the plant's physiological status.

Conclusions:

• The Bio-Master phenotyping system provides an efficient and accurate solution for comprehensive crop biochemical profiling.
• It integrates hyperspectral remote sensing with advanced modeling for improved phenotyping.
• The system surpasses conventional destructive methods in efficiency and existing nondestructive approaches in accuracy.