Roughness-driven compressive sensing AFM for accurate nanoscale surface characterization in functional material systems

• Compressive sensing hits 80 %+ accuracy in functional materials characterization. • Roughness-driven compressive sensing optimize strategy to reduce estimation error. • Compressive sensing metrics fail to guide roughness measurement strategies. Surface roughness significantly affects the functional performance of advanced materials, necessitating accurate nanoscale characterization via Atomic Force Microscopy (AFM). However, AFM’s high sampling requirements prolong measurement time and accelerate probe wear. To enhance efficiency, compressed sensing (CS) has been applied to improve AFM sampling strategies. Our literature review indicates that existing CS approaches primarily target high-quality image reconstruction, neglecting the accurate recovery of surface roughness information—a key objective in precision surface metrology. In this study, we propose a roughness-driven CS strategy to boost the efficiency and precision of AFM-based roughness measurements. We also theoretically demonstrate a nonlinear relationship between conventional CS evaluation metrics and surface roughness. Experimental results show that, compared to traditional image-based CS strategies, our method improves nanoscale roughness measurement accuracy by more than 80 %. This advancement not only reduces sampling requirements but also maintains high fidelity in roughness characterization, offering a robust tool for precise and efficient nanoscale analysis. Our approach supports performance-driven development of functional materials in fields such as microelectronics, optics, and nanomanufacturing.