A Cost-Effective and Portable Optical Sensor System to Estimate Leaf Nitrogen and Water Contents in Crops

Area of Science:

• Agricultural engineering
• Plant science
• Remote sensing

Background:

• Accurate, non-invasive measurement of leaf nitrogen and water content is crucial for plant health monitoring.
• Existing methods for determining these parameters are often costly and impractical for widespread use.

Purpose of the Study:

• To develop and evaluate a low-cost, portable multispectral sensor system for the non-invasive determination of leaf nitrogen and water content.
• To assess the sensor system's efficacy across different plant species, including canola, corn, soybean, and wheat.

Main Methods:

• A portable multispectral sensor system with visible (VIS) and near-infrared (NIR) sensors, detecting reflectance at 12 wavelengths.
• Spectral data collected from 307 leaves across nitrogen and water content experiments in a controlled greenhouse.
• Application of the rational quadratic Gaussian process regression (GPR) algorithm to correlate spectral reflectance with actual leaf content.

Main Results:

• The sensor system demonstrated strong performance in estimating leaf nitrogen content across species, with coefficients of determination (R²) ranging from 63.91% (canola) to 82.29% (soybean).
• Leaf water content estimation yielded varied results, with R² values from 18.02% (canola) to 68.41% (corn).
• The proposed sensor and regression model show promise for nitrogen determination but require further refinement for accurate water content assessment.

Conclusions:

• The developed low-cost multispectral sensor system, coupled with an appropriate regression model, is effective for non-invasively determining plant leaf nitrogen content.
• Further research and development are necessary to enhance the accuracy of leaf water content estimation using this sensor system.