Edge Detection Algorithm for Grating Images in Scanning Electron Microscopy and Its Application in Grating Uniformity Evaluation

The advancement of nanoscale metrology is critical for the development and manufacturing of advanced nanomaterials and next-generation semiconductor, where precise calibration of metrology tools using nano-grating standards is essential. However, achieving high accuracy in edge detection for these nanostructures within scanning electron microscope (SEM) and scanning probe microscope (SPM) images directly impacts measurement reliability. This study introduces a novel SEM image processing tool designed to enhance edge detection accuracy and efficiency for wafer grating standards in SEM imaging. The proposed method integrates an optimized anisotropic diffusion filter, which reduces noise while preserving structural details, with a dynamically adjusted Canny operator that adaptively calculates thresholds to improve edge localization. Experimental validation demonstrates the tool’s effectiveness: Correlation coefficients for line pitch and uniformity relative to reference measurements reach 0.93/0.96 (low magnification) and 0.99/0.98 (high magnification), respectively. These results underscore the algorithm’s capability to precisely identify nanostructure edges, significantly enhancing detection accuracy and robustness in pitch uniformity evaluation across varying magnification levels