Speaker
Description
Accurate morphological characterization of electrospun fiber mats is essential for establishing reliable structure–property relationships. Conventional manual analysis of Scanning Electron Microscopy (SEM) images is time-consuming, operator-dependent, and prone to subjective errors. Automated computational approaches offer a consistent and reproducible alternative, yet robust methods capable of simultaneously quantifying multiple structural parameters remain limited.
This study presents a computational framework for the automated morphological characterization of electrospun fibers from high-resolution SEM images. The model employs an adaptive implementation of the Sauvola thresholding algorithm for fiber network segmentation. By tuning the algorithm's sensitivity parameters, the method handles images with non-uniform contrast, enabling reliable detection of fiber boundaries and surface pores without manual intervention. Fiber diameter distributions are subsequently extracted using the Watershed algorithm, which separates overlapping or touching fibers prior to measurement. Pixel-based measurements are converted into physical units using the SEM image scale bar, enabling direct statistical analysis. Angular distributions of fiber orientations are computed from the segmented fiber skeleton. The model simultaneously quantifies four structural descriptors: fiber diameter distribution, mean fiber diameter, surface porosity, and spatial fiber orientation. Compared to conventional global thresholding methods, the adaptive Sauvola approach yields more accurate segmentation of fiber boundaries and surface pores, particularly in images with inconsistent illumination or low contrast. The automated pipeline significantly reduces processing time relative to manual inspection.
The proposed framework provides a consistent, operator-independent tool for the structural characterization of electrospun materials. By automating the extraction of key morphological parameters, it reduces human error associated with manual image analysis and offers a scalable solution for high-throughput evaluation of fiber mat quality and architecture.
| Country | Greece |
|---|---|
| Student Awards | I would like to submit this presentation into the InterPore Journal Student Paper Award. |
| Acceptance of the Terms & Conditions | Click here to agree |
