Registration Technology Of Multi-scale Measurement Data

Multi-scale 3D measurement is an important method to meet the requirements of multi-scale characterization for parts,especially in the fields of manufacture,aeronautics and cutting tools fabrication etc.The registration of multi-scale measurement data is one of the most critical technologies.Because the resolution,the information richness or the details are quite different between data at different scales,the registration can be more difficult.This thesis focuses on multi-scale processing of 3D data,characterization of scale parameters and registration of 3D data to achieve the registration of measurement data.The main research contents are as follows:(1)In order to solve the difference between the cross-scale data caused by the measurement scale,the wavelet transform is applied to the 3D data to realize the multi-scale decomposition of the original small-scale data and reduce the scale difference between the cross-scale data.(2)Aiming at the problem that the scale and difference of cross-scale data are difficult to measure and characterize,a method based on fractal dimension and local information entropy is proposed to characterize the scale parameters,which can provide an evaluation for the scale difference between multi-scale data,and provides a reference for the choice of level for wavelet transform.(3)A multi-scale data registration method based on scale decomposition is proposed.First the discrete wavelet transform is used to decompose the original data into multi-scale.Then the scale characterization technique is used to measure the scale,and finally the cross-scale data are converted to the same scale,simplifying the registration problem of cross-scale data to normal registration problem.(4)The experiment environment is built.Multi-scale measurement methods are used to obtain the 3D surface topography of the cutting tool.The 3D data are tested to verify the feasibility of the multi-scale registration method proposed in this paper.