Adaptive Compressive Sensing Imaging in AFM Based on Target Block Detection.

Area of Science:

• Surface science
• Microscopy
• Image processing

Background:

• Atomic force microscopy (AFM) is crucial for nanoscale surface analysis but traditional methods are slow.
• Compressive sensing (CS) accelerates AFM imaging, but challenges arise with increasing image size and resolution.
• Block compressive sensing (BCS) improves efficiency but struggles with consistent image quality across regions.

Purpose of the Study:

• To introduce an adaptive CS-AFM imaging scheme for faster, high-quality, high-resolution surface imaging.
• To address limitations of traditional and block CS methods in AFM.
• To improve the balance between imaging speed, quality, and resolution.

Main Methods:

• Acquire a low-resolution AFM image via fast scanning and bicubic interpolation.
• Utilize Otsu and eight-connectivity methods for target block localization.
• Employ a GRNN model to adapt sampling rates for target regions.
• Perform supplementary scans on target blocks and reconstruct using the TVAL3 algorithm.
• Replace original target regions with reconstructed high-quality blocks.

Main Results:

• The proposed adaptive CS-AFM scheme achieves significantly faster imaging speeds.
• High-quality and high-resolution images are obtained, particularly in targeted regions.
• The method demonstrates superior performance compared to existing CS-AFM techniques.
• Adaptive sampling effectively balances resolution and acquisition time.

Conclusions:

• The adaptive CS-AFM imaging scheme offers a breakthrough for rapid, high-fidelity nanoscale surface analysis.
• This innovative approach overcomes previous limitations in CS-based AFM imaging.
• The method enables efficient acquisition of detailed surface topography at micro- and nanoscales.