Theoretical And Experimental Research On Precise Measurement Of In-plane Micro Displacement In Micro Vision Image

With the development of micro-nano technology,people's perspective of exploring the laws of the world's objective things has gradually changed from macro-field to micro-field.The development of micro-nano technology is inseparable from precision measurement technology.Compared with other measurement technologies,micro-vision measurement technology has the advantages of non-contact,low cost,high precision and multi-DOF and so on.The combination of this technology and high-performance algorithms can accurately measure micro-displacement in plane.However,the existing mature image processing algorithms and displacement measurement algorithms will become inefficient after the introduction of the microscopic system and can not be directly used.The purpose of this paper is to study the in-plane micro-displacement measurement algorithm of micro-vision image with good comprehensive performance(high accuracy,strong robustness,etc).Therefore,the main research contents are as follows:(1)The adaptive enhancement method of micro-vision image is studied.Defects such as non-uniform illumination,gray value deviation,poor contrast and low signal-to-noise ratio often occur in micro-vision images,which can not be directly used as the input of micro-displacement measurement.Therefore,an adaptive enhancement algorithm of micro-vision image based on RCVG is proposed.Firstly,the overall framework of the algorithm is introduced;secondly,the process of adaptive variable gain subband decomposition is described and analyzed in detail;thirdly,illumination is estimated by using the adaptive ascending cosine convolution method based on bilateral filtering principle,and how the ascending cosine function family approximates the Gaussian function is theoretically deduced;finally,after brightness correction,RCVG algorithm is applied to the enhancement experiment of ordinary image and the pretreatment experiment of in-plane micro-displacement measurement of micro-vision image,respectively.The results show that the RCVG algorithm can effectively improve the image quality and is the precondition to ensure the measurement accuracy.(2)The micro-displacement measurement method for the translational motion of the object to be measured in the micro-vision image plane is studied.Aiming at the problem that the general measurement algorithm can not guarantee the measurement accuracy,speed and displacement measurement of any point in the micro-vision image when translation motion of the target to be measured is produced,then two methods are proposed: FTSA and OFHP.Firstly,the principle and implementation process of the two algorithms are described separately.Then,experimental research is carried out: FTSA experiments involve two different measurement systems,namely,MVMS and LIMS.The object to be measured is a 2-DOF compliant positioning stage.For OFHP experiments,MIDDLEBURY image set and some experimental image sequences are used.The results show that the proposed two measurement algorithms can achieve high measurement accuracy and good stability.In addition,OFHP can complete the displacement measurement of arbitrary pixels in the image plane.(3)The micro-displacement measurement method for the rotational motion of the object to be measured in the micro-vision image plane is studied.TWMA-RRI and THMA-RRI are proposed to solve the problems of invalidation,low accuracy and poor robustness of most common measurement algorithms when rotational motion of the target to be measured in the micro-vision image plane occurs.After introducing the measurement principle and implementation process of the two algorithms,the factors affecting the accuracy of sub-pixel algorithm,the construction of illumination model and the micro-displacement measurement using the two algorithms are simulated and analyzed,respectively.The results show that the proposed algorithm has rotation invariance,high measurement accuracy and strong robustness.Furthermore,taking the 3-DOF compliant positioning stage as the measurement object,the performances of the two algorithms are experimentally studied,which verifies the simulation results.(4)The micro-displacement measurement method for the composite motion in the microvision image plane is studied.Aiming at the composite motion of the target in the micro-vision image plane,a micro displacement measurement algorithm based on HSCT is proposed.Firstly,an improved morphological edge detection operator is introduced,which combines the sub-pixel edge detection method based on Zernike moments to get accurate edge features as the input of the algorithm.Secondly,the principle of HSCT algorithm is analyzed,including the implementation and properties of HT and HS.Then,the overall implementation of the algorithm is summarized.Finally,HS characteristic verification experiments,HSCT-based algorithm validation and accuracy verification experiments are carried out,respectively.The results show that the proposed algorithm can still guarantee high measurement accuracy when the composite motion of the target in the micro-vision image occurs.Finally,the work of this paper is summarized and the future research directions are prospected.