Feature Adaptive Sampling for Scanning Electron Microscopy.

Area of Science:

• Materials Science
• Microscopy Techniques
• Image Processing

Background:

• Scanning Electron Microscopy (SEM) is crucial for high-resolution imaging.
• Traditional SEM methods often face limitations in balancing image quality (signal-to-noise ratio, SNR) with electron dose and acquisition time.
• Improving SNR without increasing electron dose is a key challenge in SEM.

Purpose of the Study:

• To introduce a new adaptive scanning method for SEM image acquisition.
• To enhance the signal-to-noise ratio (SNR) in SEM images.
• To reduce the overall electron dose and acquisition time required for high-quality SEM imaging.

Main Methods:

• Developed an adaptive scanning technique that adjusts pixel-dwell times based on image detail.
• Implemented a two-pass scanning approach: a quick initial scan followed by targeted high-dose scanning of critical regions.
• Utilized a-posteriori image processing, including selective low-pass filtering, to combine data from both scans.

Main Results:

• Achieved approximately a 3-fold improvement in SNR compared to uniform sampling with the same acquisition time.
• Demonstrated a potential 10-fold reduction in required electron dose for adaptive scanning versus uniform scanning.
• Successfully generated a single, high-quality image by combining datasets from the adaptive two-pass scan.

Conclusions:

• The adaptive scanning method offers a significant improvement in SEM imaging efficiency.
• This technique allows for high-quality imaging with substantially reduced electron exposure and faster acquisition times.
• The developed method presents a promising advancement for various SEM applications requiring detailed analysis.