Related Experiment Video
Updated: Aug 14, 2026
05:54
Controlled Odor Mimic Permeation Systems for Olfactory Training and Field Testing
Published on: January 28, 2021
Sensor Array and SMOTE-Based Algorithms for Volatile-Fingerprint Classification of Pesticide-Treated Soil with a
Shixiao Yu1, Jiayi Li1, Hang Yu1
1Key Laboratory of Bionic Engineering, Ministry of Education, Jilin University, Changchun 130022, China.
Sensors (Basel, Switzerland)
|August 13, 2026
Summary
An electronic-nose system effectively classifies pesticide residues in soil using volatile fingerprints. This method offers rapid screening for ecological protection and food safety, overcoming limitations of traditional techniques.
Area of Science:
- Environmental Science
- Analytical Chemistry
- Sensor Technology
Background:
- Conventional pesticide-residue screening methods are time-consuming and require complex equipment.
- Rapid, large-scale screening is crucial for ecological protection, food safety, and public health.
- Limitations of current methods hinder timely detection and management of soil contamination.
Purpose of the Study:
- To develop and evaluate an electronic-nose system for classifying pesticide residues in loess soil based on volatile fingerprints.
- To assess the performance of different feature extraction and classification methods for soil pesticide analysis.
- To investigate the effectiveness of oversampling techniques for imbalanced classification of pesticide-treated soil.
Main Methods:
- A 26-sensor metal oxide semiconductor (MOS) array in a novel gas-sensing chamber was used to collect volatile fingerprints from pesticide-treated soil.
- Six common pesticides (pyrethroids, organophosphorus, organosulfur) were analyzed across four soil depths.
- Eight feature extraction methods, four classifiers, and Synthetic Minority Over-sampling Technique (SMOTE) were evaluated.
Main Results:
- The electronic-nose system demonstrated strong classification performance, particularly with k-nearest neighbors (KNN) and specific feature extraction methods.
- Synthetic Minority Over-sampling Technique (SMOTE) improved recognition of minority pesticide classes, showing consistent benefits.
- The proposed framework showed feasibility for rapid volatile-fingerprint classification of pesticide-treated soil.
Conclusions:
- The developed electronic-nose system provides a feasible approach for rapid, on-site screening of pesticide-treated soil.
- The combination of sensor arrays and SMOTE offers a promising strategy for analyzing complex soil samples.
- Further validation with larger field datasets and quantitative chemical analysis is recommended for broader application.
