Related Experiment Video
Updated: Sep 16, 2025
09:13
Implementation of a Nonlinear Microscope Based on Stimulated Raman Scattering
Published on: July 6, 2019
7.7K
Deep Learning Automated Measurements of Expanded Polystyrene Beads Size Using Low-Resolution Micrography
Alejandro E Rodríguez-Sánchez1, Héctor Plascencia-Mora2
1Facultad de Ingeniería, Universidad Panamericana, Zapopan, Jalisco, Mexico.
Microscopy Research and Technique
|July 5, 2025
Summary
This study introduces an automated Deep Learning method for measuring bead size in Expanded Polystyrene foams. The AI approach proved reliable and equivalent to manual measurements for microstructural analysis.
Area of Science:
- Materials Science
- Polymer Science
- Computational Materials Science
Background:
- Microscopic analysis of closed-cell polymeric foams, specifically bead size, is crucial for understanding material properties like thermal insulation and structural strength.
- Expanded Polystyrene (EPS) foams are widely used, and accurate bead size measurement is key to optimizing their performance.
- Traditional manual measurement methods can be time-consuming and subjective.
Purpose of the Study:
- To develop and validate an automated Deep Learning-based method for measuring bead size in Expanded Polystyrene foams.
- To compare the accuracy and reliability of the automated method against traditional manual measurements.
- To assess the suitability of the Deep Learning method for practical microstructural analysis of EPS foams.
Main Methods:
- A Deep Learning model was employed to automatically measure bead size in low-resolution micrographs of Expanded Polystyrene foams.
- Measurements were taken at two foam densities: 8.5 kg/m³ and 24 kg/m³.
- Statistical analyses, including Student's t-test, Levene's test, Mann-Whitney U test, and Bland-Altman plots, were used for comparison.
Main Results:
- No significant differences were found between the automated Deep Learning method and manual measurements for bead size.
- Student's t-test and Levene's test confirmed comparable means and variances between the two methods.
- Bland-Altman analysis showed no systematic bias, indicating equivalence and reliability of the automated approach.
Conclusions:
- The proposed Deep Learning-based method is a reliable and precise alternative to manual bead size measurement in Expanded Polystyrene foams.
- This automated approach is suitable for practical, large-scale microstructural analysis of EPS materials.
- The findings support the adoption of AI-driven techniques for material characterization, enhancing efficiency and accuracy.
