Two-in-One Implementation of Noise Reduction and Incline Emendation for Atomic Force Microscopic Images

Abstract

Apart from random noise, an atomic force microscopic (AFM) image may also suffer from an inclined plane owing to the in-coincidence between the scanning plane and the sample plane. While the wavelet-shrinkage-based image denoising technique was applied to remove the noise in the AFM image, the least-square-fitting technique was applied to the approximate sub-image so as to emendate the inclined plane. By using this method, the image de-noising and incline emendation can be implemented in parallel. As incline emendation was implemented in the approximate sub-band, the time cost for incline fitting is greatly reduced, and hence the real-time performance of the algorithm is improved. In addition, as the algorithm was implemented by means of wavelet transform, it can be easily combined with the wavelet-based image compression, which is beneficial to in-situ analysis and processing of AFM images. Through experiments, the new method was evaluated by employing an AFM image and the results were given and discussed.