Research On Optical Measurement Of Micronano Scale Surface Topography Based On Focus Variation Microscopy

Surface topography is a vital geometric feature,especially for engineering parts.By controlling the surface topography of a component,not only the mechanical performance can be affected,but its property and function can also be fundamentally altered.Therefore,high-precision measurement of surface topography has wide application in both industrial and academic field,and it’s of great research value and prospect.This thesis researches on the key techniques of focus variation microscopy,including focus measurement,depth estimation,and point cloud optimization,etc.Experimental software and hardware systems are also developed.The main contents of research are as follows:(1)In order to handle the difference between actual focus curve and ideal Gaussian curve,a focus curve fitting method based on adaptive interval is proposed,which can utilize as many valid data as possible while keeping the computational complexity low;(2)As there are lots of noise points(points with error depth)in the point cloud,this thesis studies on the distribution regularity of noise points and proposes a noise reduction method base on median filter,which can improve the quality of the point cloud;(3)In order to estimate the depth of some points which is hard to calculate directly due to lack of surface texture,an approach based on spatial consistency model is proposed for point cloud optimization.Moreover,this thesis combines the advantages of both small window and large window,and improves local precision of the point cloud while keeping its mean error low;(4)By studying the basic principle of optical imaging,a defocused image simulation system is developed based on the point spread model.A method for calculating the spread parameter between infocus and out-of-focus images is also proposed.The simulation system can simulate vertical scanning procedure carried out by real hardware system,and synthesize sequential images with limited depth of focus,which helps a lot when analyzing the problems about 3D reconstruction algorithms;(5)A prototype based on research results above is implemented,including software and hardware systems.By carrying out actual measurement experiments,the feasibility of proposed methods is verified.