Research On The 3D Measurement Technology Of Digital Microscopy

In recent years,with the development of science and technology,the object of bio-medical,precision manufacturing,materials science and industrial micro-electromechanical systems fields gradually transited from the macro field to micro field.It also requires the corresponding measurement developed to micro measurement.As a classic microscopic instrument,optical microscopy is mainly used for human observation of microscopic objects.It's shortcomings of complex operation procedure,low degree of automation,low resolution and high cost problems make it is unable to meet the requirements of modern industry.Therefore,it is of practical significance to research the measurement technology based on optical microscopy,especially three dimensional(3D)measurement technology,to realize low cost,high resolution and intelligent micro measurement system.To achieve this goal,this paper focuses on the following aspects of research and exploration:(1)3D measurement technology based on stereo light microscopy: Stereo light microscope with narrow vision and shallow depth of field is widely used in micro-domain research.In this paper,a depth estimation method of micro objects based on geometric transformation is proposed.Through the analysis of optical imaging geometry,the definition of geometric transformation distance is given and the depth-distance relation express is derived.The parameters of geometric transformation and express are calibrated with calibration board images captured in aid of precise motorized stage.The depth of micro object can be estimated by calculating geometric transformation distance.The proposed depth-distance relation express is verified in the experiment and depth map of an Olanzapine tablet surface is reconstructed.For light microscopy images have the characteristics of shallow depth of field,serious distortion and poor resolution,mismatch is a ubiquitous phenomenon.The paper presents a mismatch detection method for stereo light microscopy stereo matching.Affine transformation matrix and match constraint condition are calibrated with precision solid dots calibration board on a motorized stage.Affine transformation bias vector of each match pair is taken as criteria to apply mismatch detection.The experimental results show that the method can detect more mismatch pairs and preserve more match pairs than traditional method.(2)Microscopic image clarity: Besides how to eliminate the influence of image clarity,we can make use of image clarity information.In respect of passive elimination,image deconvolutionis an effective image restoration technique to improve the quality of digital microscopic images resulting from out-of-focus blur.To solving the severely ill-posed problem of traditional Richardson-Lucy method,a new microscope image restoration method is proposed based on multiple defocused images of different aperture and considering the point spread difference of various directions.The maximum-likelihood estimation is used to suppress the ringing artifacts and noises sensitivity of microscope image.Experimental results show that the proposed algorithm performs better than Richardson–Lucy method and improves peak-signal-to-noise-rate about 4dB.In respect of active mode,for optical microscopy image is often out of focus,a 3D measurement method based on focus measure is proposed.First,in order to make the same evaluation object,dot centers of different depth calibration board are extracted with region grown segmentation algorithm and regions to be evaluated are selected.Second focus measure function is used to evaluate the selected image regions and definition-depth curve is obtained.Then the curve peak is got by curve fitting technique,the depth is draw and 3D depth measurement is completed.Finally coplanarity of 3D dot on calibration board and plane angle measured is used to evaluate the measurement results.Experimental results show than the proposed method can measure 3D depth well.The method can be used in tilt measurement and automatic leveling in microscopic system.(3)A microscopic 3D measurement method based on computational photography: To break the limitation of the depth of field of an optical microscope image,a 3D measurement method with super depth of field is proposed.Light field information of different angles is obtained by moving aperture and the 3D scene is reconstructed by computational reconstitution method.Firstly,stereo matching of different aperture position images is carried out to obtain the multi aperture imaging deviation,and the focus plane moving distance is estimated.Then,the relation expression between the image coordinates and the focus plane moving distance is determined according to the image coordinate.Then the 2D coordinate of space point is obtained by the expression coefficients.Finally,the depth coordinates are computed,and the 3D reconstruction of the spatial points is completed.3D measurement of the calibration board with different angles and circuit board are carried out.The experimental results show that the maximum error of distance measurement is0.84%,and the maximum angle measurement error is 4.56%.In this paper,aim at the problems existing in the process of optical microscopy,image processing and 3D measurement,a series of research work are carried out and a number of relatively complete 3D measurement systems are constructed.The research contents include image preprocessing,the imaging model building,system calibration,stereo matching,depth from defocus and computational photography.these are the difficult problems in the process of microscopic three-dimensional measurement.It has certain application value and innovation.