Related Experiment Video
Updated: Jul 6, 2025
10:16
Protocol for Microplastics Sampling on the Sea Surface and Sample Analysis
Published on: December 16, 2016
49.7K
Deep convolutional neural networks for aged microplastics identification by Fourier transform infrared spectra
Ganning Zeng1, Yuan Ma2, Mingming Du2
1College of Environment, Zhejiang University of Technology, Hangzhou 310014, China; Key Laboratory of Ocean Space Resource Management Technology, MNR, Hangzhou 310012, China.
The Science of the Total Environment
|December 30, 2023
Summary
A new deep convolutional neural network (CNN) model accurately identifies aged microplastics (MPs) using infrared spectra. This advanced technique overcomes challenges in MP classification, achieving 100% accuracy.
Area of Science:
- Environmental Science
- Analytical Chemistry
- Materials Science
Background:
- Microplastic (MP) detection is crucial for environmental monitoring.
- Aging processes complicate the accurate identification and classification of MPs.
- Infrared (IR) spectroscopy is a key technique for MP analysis.
Purpose of the Study:
- To develop a robust classification model for aged microplastics (MPs) using infrared (IR) spectra.
- To enhance the accuracy and efficiency of MP identification in environmental samples.
- To overcome the limitations of traditional methods in classifying aged MPs.
Main Methods:
- A deep convolutional neural network (CNN) model was developed using original IR spectra.
- The model incorporated Adam optimization, Dropout for regularization, and ReLU activation.
- Performance was evaluated against Artificial Neural Network, Random Forest, and Deep Neural Network models.
Main Results:
- The CNN model achieved 100% accuracy in identifying aged microplastic samples.
- CNN demonstrated superior feature extraction and recognition capabilities compared to other methods.
- The model simplified the pre-processing procedure for spectral data.
Conclusions:
- Deep convolutional neural networks offer a powerful and accurate solution for classifying aged microplastics.
- The developed CNN model is versatile, suitable for limited data, and expandable for future applications.
- This approach significantly advances the field of microplastic identification and environmental monitoring.
