Research On Measurement Method Of Microstructure Feature Size Based On Scanning Electron Microscope

The structural characteristics of micro components are important information for functional design and manufacturing,and are closely related to the service performance of products.With the rapid development of modern micro and nano manufacturing technology,micro components are gradually becoming finer,more precise and more complex materials,which makes the crossscale feature size measurement of micro structures face more urgent and demanding needs and challenges.Although Scanning Electron Microscope(SEM)is widely used for cross-scale feature size measurement and analysis of microstructures due to its advantages of high accuracy,non-contact and wide measurement range,it is subject to interference from internal and external factors,which causes degradation of image quality in practical applications.To address the urgent need for cross-scale feature size measurement of microstructures and the challenges faced by SEM measurement techniques,the following studies are conducted in this paper:(1)An SEM-based micro-nano operation and measurement system is built for the demand of cross-scale feature size characterization and analysis of super-field microstructures.In order to solve the stability analysis of the system and the calibration problem of cross-scale large range variable magnification,a fast sub-pixel template matching method based on Log search is firstly proposed for the static and dynamic stability analysis of the system;and a calibration model and parameter solution method with continuously variable magnification is further proposed to realize the de-calibrated plate high precision calibration of the system by using the motion characteristics of the nanopositioning stage.Finally,the proposed method is proved to be effective in ensuring the accuracy and reliability of system calibration and measurement through experiments,which provides a reliable experimental basis for the subsequent research of related methods.(2)The noise reduction and contrast and visualization enhancement methods of SEM images are studied.To address the degradation of image quality caused by the combination of internal and external coupling factors such as microstructure morphological features and material properties,the system itself and the surrounding environment,a partial differential equationbased microstructure image fusion contrast and pseudo-color visualization enhancement method is proposed by investigating the noise sources and mechanisms of SEM images.Both simulations and experimental cases show that the proposed method can effectively suppress and eliminate local shadows and noise in SEM images,and the enhanced images have clearer and more discernible microstructure morphology information,which provides reliable data guarantee and technical support for the subsequent measurement and analysis of microstructure features.(3)A precision measurement method for the size of local features of microstructures in the SEM field of view is investigated.In order to solve the edge detection problem of microstructure local features,firstly,a gray-scale histogram-based edge highlight detection and correction method is proposed to suppress and eliminate the edge effects of microstructure images;and further,an adaptive partial differential equation diffusion-based method is proposed to reduce the effects of pseudo-edges and residual noise.Finally,in order to realize the precise measurement of the local feature size of microstructures,a line scan contour-based feature edge accurate positioning and measurement method is proposed.Both simulation and experimental results show that the proposed method has strong robustness and stable and reliable measurement accuracy,which can effectively guarantee the implementation of precision measurement and analysis tasks of local feature dimensions of microstructures.(4)A precision measurement method for the global feature size of SEM super-field microstructures is investigated.In order to reach global high-resolution imaging of super-field microstructures,a fast image stitching method based on gray-level equalization and adaptive fading-in and fading-out fusion is proposed.The method effectively solves the problems of detail redundancy,noise interference and grayscale perturbation of the imaging system in SEM images.Based on the obtained super-field high-resolution images,a method based on Zernike sub-pixel edge detection is proposed to realize the precision measurement of cross-scale feature size of microstructures.Comprehensive experiments demonstrate that the proposed method can effectively achieve global high-resolution imaging of microstructures in super-field and precise measurement and analysis of cross-scale feature size.Finally,the main works and innovations of this thesis are summarized,and the future research directions are prospected.