Using machine learning techniques to extract nanomaterial information from SEM images

Quantitative analysis of scanning electron microscopy (SEM) images is important for understanding the structural and morphological variations of nanomaterials. Information obtained from SEM images—such as the particle size, size distribution, and morphology—provide technical and scientific insights into the process of nanomaterial synthesis, fabrication, and manufacturing. Previous software aimed at extracting relevant information from SEM images required time-consuming manual intervention and struggled to measure diameters of overlapping nanoparticles. In addition, the methods were generally limited to specific applications.

To address these limitations, Lawrence Livermore researchers used computer vision and machine learning techniques to develop a generally applicable approach to quantitatively extract particle size, size distribution, and morphology information from SEM images. The proposed approach employs computer vision and machine learning techniques and offers fully automated, high-throughput measurements with little user intervention, even when overlapping nanoparticles, rod shapes, and core–shell nanostructures are present.

The team demonstrated the effectiveness of the proposed approach by performing experiments on SEM images of nanoscale materials and structures with different shapes and sizes. The proposed approach shows promising results when compared with manually measured sizes. The algorithm and its implemented GUI software package, named Livermore SEM Image Tools (LIST), are publicly available as an open source software code, and the code can easily be adopted to different image analysis tasks.

This research received support from the Laboratory Directed Research and Development Program (16-ERD-019 and 19-SI-001).

[H. Kim, J. Han, and T.Y. Han, Machine vision-driven automatic recognition of particle size and morphology in SEM images, Nanoscale 12, 19461 (2020), doi: 10.1039/D0NR04140H.]