Related Experiment Video
Updated: Jun 28, 2025
10:16
Protocol for Microplastics Sampling on the Sea Surface and Sample Analysis
Published on: December 16, 2016
49.6K
Fast Detection and Classification of Microplastics below 10 μm Using CNN with Raman Spectroscopy
Jeonghyun Lim1, Gogyun Shin1, Dongha Shin1
1Department of Chemistry and Chemical Engineering, Inha University, Incheon 22212, Republic of Korea.
Analytical Chemistry
|April 16, 2024
Summary
Artificial intelligence (AI) accelerates microplastic (MP) detection using convolutional neural networks (CNNs). This method rapidly classifies plastic types in small particles, aiding global microplastic mapping efforts.
Area of Science:
- Environmental Science
- Analytical Chemistry
- Data Science
Background:
- Microplastic (MP) pollution is a global environmental concern.
- Raman spectroscopy is effective for MP detection (<10 μm) but suffers from low signal intensity, leading to long acquisition times.
- Fourier transform infrared (FTIR) spectroscopy has diffraction limitations for smaller particles.
Purpose of the Study:
- To develop a faster method for identifying microplastics (MPs) using AI.
- To overcome the limitations of Raman spectroscopy for MP detection.
- To enable rapid collection of microplastic distribution data.
Main Methods:
- Implementation of a convolutional neural network (CNN) model.
- Development of a tailored data interpolation strategy.
- Application to microplastic particles in the 1-10 μm size range.
Main Results:
- Achieved classification of plastic types for individual microplastic particles.
- Reduced exposure time to just 0.4 seconds per particle.
- Reached an approximate confidence level of 85.47(±5.00)%.
Conclusions:
- The AI-driven approach significantly accelerates microplastic identification.
- This method facilitates faster aggregation of microplastic distribution data.
- Contributes to the development of a global microplastic map.
