Analysis method for individual particle shape and chemical composition classification of atmospheric samples using SEM-EDS automatic analyzer
The use of a platinum-coated polycarbonate filter with SEM-EDS and AI training addresses the challenges of analyzing ultrafine dust, enabling efficient classification and source identification of fine dust particles.
Patent Information
- Application Number
- KR1020240195564
- Authority / Receiving Office
- KR · KR
- Patent Type
- Patents
- Current Assignee / Owner
- Filing Date
- 2024-12-24
- Publication Date
- 2026-07-21
- Estimated Expiration
- 2044-12-24
AI Technical Summary
Conventional methods for analyzing individual fine dust particles are labor-intensive and difficult to apply universally for statistical analysis, especially for ultrafine dust like PM2.5, and existing SEM-EDS analysis faces challenges with fibrous filters and diverse compound compositions.
A method involving a polycarbonate filter coated with platinum for capturing atmospheric samples, combined with SEM-EDS analysis and artificial intelligence training, to classify particle shapes and compositions, using unsupervised learning and manifold learning to identify trends.
Enables efficient classification of particle sources by distinguishing contamination sources through automated analysis, reducing the time required to identify pollution origins and providing statistical analysis of large quantities of particles.
Smart Images

Figure 112024143596187-PAT00020_ABST
Abstract
Citation Information
Patent Citations
Powder dust measurement device and powder dust measurement method
JP2020060386A
Test strip, and microorganism sensor device and sensing method
KR1020220037760A
Method for automatically analyzing micro particles distributed over a large area using artificial intelligence, electron microscope and EDS analyzer
KR1020230063276A
Apparatus and method for fine dust sensor calibration based on machine learning
KR102504412B1
Particle manipulation system with camera / classifier confirmation and deep learning algorithm
US11898954B2