Research On Key Technology Of Image Processing For Scanning Probe Microscope

Scanning Probe Microscopy(SPM)can achieve the material's nano-scale topography imaging.While it's a precise and sophisticated system with the unity of light,machine and electrical technologies,it is inevitable to involve many defects in the imaging process.In order to carry out material science experiments more effective,common defects need to be processed to obtain high-quality topography images.In this paper,with the processing of SPM image defects as the focus,the causes of common defects are analyzed in detail,and better treatment methods are proposed for each image defect.The contents of this article are as follow:(1)A B-spline surface fitting algorithm with automatic generation mask is constructed.Firstly,the edge of the sample is identified and masked by the edge detection and threshold analysis.Then B-spline surface fitting is used to fit the background of the non-step region.Finally,the background is subtracted from the original topography.The outcome is obviously superior to the traditional method.(2)The traditional Tip Blind Reconstruction algorithm is improved,then with the reconstructed tip as the degenerate function,the constrained least squares filtering algorithm is used for the deconvolution of the degenerate and fuzzy topography image,and the recovery outcome is better than the Lucy-Richardson blind deconvolution's.(3)The traditional Criminisi texture restoration algorithm is optimized,then the TV model repair algorithm and Criminisi algorithm are used to deal with the structure type and texture type damage in the topography image,and terrific image repair outcome can be obtained.(4)A frequency localization notch filtering algorithm is constructed.Firstly,a two-dimensional discrete Fourier Spectrum of the noisy image is obtained.And then with the help of Sobel edge detector the noise frequency point is located accurately,based on which a general oval notch filter with adjustable long diameter,short diameter and order is created.This method works well with various types of periodic noise.(5)A topography image processing software is realized by using C#and MATLAB mixed programming.In the software,almost all types of topography image processing algorithms are realized,such as background subtraction,degradation fuzzy recovery,damage repair,periodic noise denoising,contrast enhancement,resolution enhancement and signal enhancement.