Reconstruction of an AFM image based on estimation of the tip shape

From the viewpoint of mathematical morphology, an atomic force microscopy (AFM) image contains the distortion effect of the tip convolution on a real sample surface. If tip shape can be characterized accurately, mathematical deconvolution can be applied to reduce the distortion to obtain more precise AFM images. AFM image reconstruction has practical significance in nanoscale observation and manipulation technology. Among recent tip modeling algorithms, the blind tip evaluation algorithm based on mathematical morphology is widely used. However, it takes considerable computing time, and the noise threshold is hard to optimize. To tackle these problems, a new blind modeling method is proposed in this paper to accelerate the computation of the algorithm and realize the optimum threshold estimation to build a precise tip model. The simulation verifies the efficiency of the new algorithm by comparing the computing time with the original one. The calculated tip shape is also validated by comparison with the SEM image of the tip. Finally, the reconstruction of a carbon nanotube image based on the precise tip model illustrates the feasibility and validity of the proposed algorithm