Super-resolution AFM imaging based on compressive sensing

Abstract

Atomic force microscopy (AFM) is a powerful and ultra-precision instrument in nano-scale, which is widely used in many fields. It is a complex and time-consuming process for AFM imaging. Most of the original AFM images are with low resolution. For nano-scale measurement and imaging, it is very important to obtain super-resolution images. In most cases, super-resolution imaging takes a long time and the quality of imaging is unsatisfactory. In this regard, a novel super-resolution imaging method based on compressed sensing (CS) technology is proposed in AFM. In the experiment, six samples with different morphology were used to test the effect of super-resolution image reconstruction with different upscaling factors (2, 3 and 4). The quality of reconstructed image is analyzed and evaluated by image evaluation metrics (PSNR and SSIM). The relationship between the reconstruction quality of different images and the actual TV or TV/R of sample images is analyzed, which can provide a preliminary basis for predicting the imaging quality. Comparing with other super-resolution imaging methods, the proposed method has achieved better imaging effect both visually and quantitatively. In summary, super-resolution imaging method based on CS not only has high imaging quality but also has high speed.