Related Experiment Video
Updated: Sep 17, 2025
11:37
RGB and Spectral Root Imaging for Plant Phenotyping and Physiological Research: Experimental Setup and Imaging Protocols
Published on: August 8, 2017
16.4K
Hyperspectral quantitative retrieving of soil iron oxide and Zn content combining feature selection and machine
Jiankai Hu1, Lin Hu2, Shu Gan2
1Faculty of land and Resources Engineering, Kunming University of Science and Technology, Kunming 650093, China.
Summary
Hyperspectral reflectance effectively estimates soil iron oxide and zinc content using spectral transformations, feature selection, and machine learning. Optimal models vary, with FD_CARS_SVM for iron oxide and FD_Boruta_XGBoost for zinc, demonstrating improved accuracy.
Area of Science:
- Soil Science
- Remote Sensing
- Analytical Chemistry
Background:
- Accurate estimation of soil physicochemical properties is crucial for environmental monitoring and agricultural management.
- Hyperspectral reflectance offers a non-destructive method for assessing soil composition, including iron oxide and heavy metal content.
- Optimizing spectral preprocessing, feature selection, and machine learning models is key to improving retrieval accuracy.
Purpose of the Study:
- To evaluate the effectiveness of various spectral transformations and feature selection methods for hyperspectral data.
- To identify the optimal machine learning model combination for estimating soil iron oxide and zinc (Zn) content.
- To compare the performance of different methods in retrieving soil physicochemical properties.
Main Methods:
- Applied spectral transformations: Continuum Removal (CR), Standard Normal Variate (SNV), First Derivative (FD), and Second Derivative (SD).
- Utilized feature selection algorithms: Competitive Adaptive Reweighted Sampling (CARS) and Boruta.
- Constructed and evaluated machine learning models: Partial Least Squares Regression (PLSR), Support Vector Machine (SVM), Back Propagation Neural Network (BPNN), and Extreme Gradient Boosting (XGBoost).
Main Results:
- Spectral transformations effectively reduced external interference and highlighted spectral features, enhancing band selection and model prediction accuracy.
- CARS proved more suitable for soil iron oxide selection, while Boruta was better for heavy metal Zn.
- The optimal model for iron oxide was FD_CARS_SVM (R²C=0.878, R²V=0.849), and for zinc, it was FD_Boruta_XGBoost (R²C=0.999, R²V=0.682).
Conclusions:
- Spectral transformations significantly improve the accuracy of estimating soil iron oxide and zinc content using hyperspectral data.
- The choice of feature selection method is critical and depends on the target soil property (iron oxide vs. zinc).
- Different machine learning models show varying suitability for linear (iron oxide) and nonlinear (zinc) relationships between spectral data and soil properties.
Related Concept Videos
02:35Key Elements for Plant Nutrition
21.8K
Like all living organisms, plants require organic and inorganic nutrients to survive, reproduce, grow and maintain homeostasis. To identify nutrients that are essential for plant functioning, researchers have leveraged a technique called hydroponics. In hydroponic culture systems, plants are grown—without soil—in water-based solutions containing nutrients. At least 17 nutrients have been identified as essential elements required by plants. Plants acquire these elements from the...
21.8K
01:17Precipitation and Co-precipitation
2.1K
Precipitation and coprecipitation methods can be used to separate a mixture of ions in a solution. In qualitative inorganic analysis, ions that form sparingly soluble precipitates with the same reagent are separated based on the differences in solubility products. For example, consider the separation of Cu(II) and Fe(II) ions by precipitation as insoluble sulfides. First, copper(II) sulfide is precipitated by the addition of acidic H2S, where the dissociation of H2S is suppressed. Adding H2S...
2.1K