Related Experiment Video
Updated: Oct 31, 2025
08:19
Lensless On-chip Imaging of Cells Provides a New Tool for High-throughput Cell-Biology and Medical Diagnostics
Published on: December 14, 2009
12.2K
Low-cost smartphone-based LIBS combined with deep learning image processing for accurate lithology recognition
Xu Wang1, Sha Chen1, Mengfan Wu2
1Research Centre of Analytical Instrumentation, School of Mechanical Engineering, Sichuan University, Chengdu, China 610065, P. R. China. yduan@scu.edu.cn.
Chemical Communications (Cambridge, England)
|June 29, 2021
Summary
A novel smartphone spectrometer for laser-induced breakdown spectroscopy (LIBS) enables simultaneous detection of atomic data and plasma plume images. This allows for accurate rock type prediction using deep learning models directly on raw spectral data.
Area of Science:
- Analytical Chemistry
- Spectroscopy
- Geoscience
Background:
- Laser-induced breakdown spectroscopy (LIBS) is a powerful analytical technique.
- Developing portable and cost-effective LIBS systems is crucial for field applications.
- Existing LIBS systems often lack the capability to capture spatial and temporal plasma plume information.
Purpose of the Study:
- To develop a low-cost, multi-channel smartphone-based spectrometer for LIBS.
- To integrate simultaneous detection of atomic emission and spectral imaging of laser-induced plasma plumes.
- To enable accurate rock type prediction using advanced data processing techniques.
Main Methods:
- A multi-channel spectrometer was designed and coupled with a smartphone.
- A linear array of optical fibers collected light from the laser-induced plasma.
- A 2D CMOS detector captured both spectral and spatial information of the plasma plume.
- A deep learning model was employed to process the raw spectral and image data.
Main Results:
- The developed spectrometer successfully achieved simultaneous multi-channel detection.
- Spectral images revealing plasma plume propagation and spatial distribution were recorded.
- Accurate rock type prediction was achieved by directly processing raw data with a deep learning model.
- The system demonstrated a low-cost and portable solution for LIBS analysis.
Conclusions:
- Smartphone-based LIBS spectrometers can be effectively developed for field analysis.
- Integrating spectral imaging with atomic emission analysis enhances LIBS capabilities.
- Deep learning models can directly process complex LIBS data for accurate material identification.
