An automated method for large-scale SEM image acquisition and analysis for metasurface materials

Abstract

Scanning electron microscopy (SEM) is an essential tool for characterizing micro/nano structures of metasurfaces. However, in large-scale sample analysis, challenges remain, including low imaging efficiency, insufficient stitching accuracy, and heavy reliance on manual operations for structural parameter extraction. To address these issues, this paper proposes an automated large-scale SEM image stitching and analysis method tailored for metasurface structures. The method acquires image sequences with overlapping regions through an automated acquisition strategy and achieves high-precision image reconstruction by combining sequence optimization with multi-level stitching techniques. Based on the stitched results, automatic extraction and statistical analysis of structural unit cells are further performed. Experimental results demonstrate that the proposed method enables stable stitching of large-scale SEM images and effectively supports automated characterization and analysis of metasurface structures.