An Effective Correction Method for AFM Image Distortion due to Hysteresis and Thermal Drift

Abstract

Hysteresis and thermal drift cause image distortion of atomic force microscopy (AFM). To address this issue, a hysteresis compensation algorithm based on B-spline curve fitting is first designed to correct the image distortion due to hysteresis, so as to provide a basis for further thermal drift correction. Afterward, a novel off-line drift correction algorithm based on cross diagonals' scanning is proposed to reconstruct high-quality AFM images. More precisely, according to the distorted image and the leading diagonal profile obtained by initial scanning, the drift is first calculated by template matching with an increasing sliding window. Furthermore, since the calculated drift contains a part of incorrect data, an adaptive filter is designed to generate more accurate horizontal thermal drift, based on which an undistorted image is constructed by bilinear interpolation. Finally, the performance of the proposed method is verified by convincing experimental and application results. Compared with online correction methods, the proposed approach is easier to implement since it does not require hardware improvement. Besides, compared with other off-line methods, the proposed method is simpler with less time consumption since it only needs two additional scanning lines to correct image distortion.