Related Experiment Video
Updated: Apr 28, 2026
11:37
RGB and Spectral Root Imaging for Plant Phenotyping and Physiological Research: Experimental Setup and Imaging Protocols
Published on: August 8, 2017
17.0K
From leaf to canopy: Inversion of lettuce pigment distribution using hyperspectral imaging technology combined with
Yue Zhao1,2,3, Jiangchuan Fan2,3, Xianju Lu2,3
1College of Information and Electrical Engineering, China Agricultural University, Beijing, 100083, China.
Plant Phenomics (Washington, D.C.)
|April 27, 2026
Summary
This study introduces a non-destructive method using hyperspectral imaging and deep learning (LPCNet) to accurately assess lettuce pigment content. The approach enables high-throughput analysis and visualization of pigment distribution for precision agriculture.
Area of Science:
- Agricultural Science
- Plant Physiology
- Remote Sensing
Background:
- Plant pigment content is vital for photosynthetic efficiency and plant health.
- Traditional pigment measurement methods are destructive, inefficient, and costly for precision agriculture.
- High-throughput, non-destructive methods are needed for spatial analysis of plant pigments.
Purpose of the Study:
- To develop a cross-scale, non-destructive method for detecting lettuce pigment content.
- To integrate hyperspectral imaging (HSI) with deep learning for accurate pigment prediction.
- To visualize the spatial distribution of pigments from leaf to canopy.
Main Methods:
- A multidimensional dataset was created using eight lettuce varieties.
- A deep learning model, LPCNet, was developed, integrating CNN, BiLSTM, and MHSA mechanisms.
- HSI data was combined with a leaf-level inversion model for canopy-level visualization.
Main Results:
- LPCNet achieved high predictive accuracy for chlorophyll a (R²=0.9449), chlorophyll b (R²=0.8613), carotenoids (R²=0.9121), and total pigment content (R²=0.8476).
- The model simplified feature extraction compared to traditional machine learning.
- Spatial distribution of pigment content across lettuce canopies was successfully visualized.
Conclusions:
- The proposed HSI and deep learning approach offers a rapid, accurate, and non-destructive method for assessing lettuce pigment content.
- LPCNet provides an effective visualization tool for understanding lettuce physiological status and development.
- This method supports advancements in precision agriculture through high-throughput spatial analysis.
