Research On Atomic Force Microscope Image Reconstruction Based On Lucy-Richardson Algorithm

With the research such as semiconductor,nano-materials,biology,chemical industry,food and medicine has reached the level of nanoscale,(Atomic Force Microscope,AFM)as an important tool for nano-detection and nano-fabrication has gained a large number of practical applications in the field of nanotechnology because of its features,such as high resolution,good repeatability and controllability.The mechanism of action that AFM observe and control sample is using the probe tip as end actuator to perform high resolution and high precision observation operation of interest region,by measuring the interaction force between the probe and the sample surface.Because there are some uncertain factors in AFM scanning imagery,such as probe drive PZT nonlinear,system drift,the probe tip broadening effect and other factors,resulting in AFM images accuracy are less than ideal requirement,the scanned images and the samples' topography are not completely consistent.The AFM images accuracy seriously restricted the AFM applications in the field of nano-operation.The interaction between the probe tip and the sample surface creates AFM images,consequently,the uncertainty of Atomic Force Microscope measuring mostly results from the uncertainty of the probe tip shape.So,this thesis reasarch AFM images reconstruction on AFM images distortion caused by probe tip shape and probe tip broadening effect,mainly including the follwing aspects.For the issue of imprecision on probe tip shape,this thesis researches on probe model reconstruction using blind modeling algorithm based on mathematical morphology;and because of the disadvantages of blind modeling algorithm,such as complex calculation,long operation time and low efficiency,this thesis proposes a probe blind modeling algorithm which can increase algorithm speed and precision to establish probe tip shape,so that the algorithm can be applied to the correction and reconstruction of AFM images.For the issue of probe tip broadening effect,an algorithm of AFM image reconstruction based on Lucy-Richardson(L-R)is proposed to improve the scanned images expansion problems caused by probe tip and for the issue of slow convergence speed and amplifying noise existing in L-R algorithm,a new algorithm named accelerating regularization L-R recovery algorithm is proposed.The new algorithm increases its convergence speed by vector extrapolation and handles noise by using regularization to handle noise residuals formed in the process of iteration of original L-R algorithm to make up the deficiency of the original L-R algorithm.The experiments prove that the nanoparticle morphology after reconstruction is closer to the nanoparticle morphology parameters provided by the manufacturer,the AFM probe tip broadening effect is effectively reduced,and the AFM images after reconstruction have better image accuracy.This makes the AFM images have a value of practical application.