Adaptive under-sampling strategy for fast imaging in compressive sensing-based atomic force microscopy.

Area of Science:

• Atomic Force Microscopy
• Computational Imaging
• Materials Science

Background:

• Compressive sensing (CS) simplifies atomic force microscope (AFM) scanning, enabling high-quality imaging.
• Random under-sampling in CS-AFM often results in detail loss in regions of interest (ROIs), potentially missing crucial microscopic phenomena.

Purpose of the Study:

• To develop an adaptive under-sampling strategy for CS-AFM to optimize sampling processes.
• To improve image quality and imaging speed in CS-AFM, particularly in ROIs.

Main Methods:

• Implemented an adaptive under-sampling strategy by assigning weight coefficients to ROIs and regions of base (ROBs).
• Controlled the distribution of under-sampling points and the measurement matrix based on these weights.
• Validated the method through extensive simulations and experiments on a custom-built AFM.

Main Results:

• Demonstrated a clear relationship between ROI weight coefficients and resulting image quality in simulations.
• Successfully verified the effectiveness of the adaptive strategy on experimental AFM data.
• Showcased optimized sampling that significantly enhances image quality and accelerates AFM imaging.

Conclusions:

• The proposed adaptive under-sampling strategy effectively optimizes the CS-AFM sampling process.
• This method leads to improved image quality and faster imaging speeds without compromising essential details.
• The technique offers a significant advancement for high-quality, efficient atomic force microscopy applications.