Industrial CT Image Measurement Method And Application For Precision Component

In recent years,the internal dimension measurement of precision component with high precision by Computed Tomography(CT)has gradually developed into an important technical means for product quality control.In practical engineering applications,when measuring geometrical dimensions and other parameters of the workpiece through CT images,the accuracy of image segmentation and measurement is seriously affected by the existence of different degrees of artifacts in the reconstructed images,which are caused by differences in the materials and structures of the workpieces to be inspected and the performance of CT equipment such as radiation energy.Therefore,how to improve the image quality and the high-precision measurement of industrial CT images of precision component has always been one of the difficulties and focuses in industrial CT research.Aiming at this problem,this thesis carried out the following research works:First,the advantages and disadvantages of three classical enhancement algorithms,namely guided filtering,structural tensor combined with non-local mean and wavelet transform(WT)were analyzed.The related algorithms were programmed and compared.The actual CT image enhancement results show that wavelet transform has obvious advantages in improving the images quality,which can effectively reduce the influence of artifacts and better enhance the image details.Second,a subpixel image segmentation method combining wavelet transform and minimization of region-scalable fitting energy model(RSF)was proposed,denote as the WT-RSF method.In order to achieve the high precision measurement of the geometric elements of precision component,the primary challenge was to achieve the accurate segmentation of CT images of precision component.Aiming at the problem of low contrast and weak edge of industrial CT images of precision component,The WT-RSF method achieved image enhancement and accurate segmentation by combining the advantages of good local characteristics of wavelet transform and high-precision edge positioning of RSF model,thus laying a good foundation for the measurement of precision component at a later stage.Meanwhile,the segmentation accuracy of WT-RSF method and Chan-Vese(CV)method was compared and analyzed through actual experiments.The experimental results show that the WT-RSF method can effectively achieve the segmentation of weak edge CT images of the turbine blade,and the relative errors of the actual measurement results of the gauge block group are less than 0.7%,which is 1.4 times higher than the measurement accuracy of CV method.Third,WT-RSF method was used to measure the geometric elements of precision component.This chapter focused on the realization of the measurement and verification of geometric elements such as diameter,angle and wall thickness.The images were segmented by WT-RSF method,and then the data points at the edges of the region of interest were extracted.When the equation of the circle or line was obtained by the least square fitting method,the value of the geometric element to be measured could be determined.To measure the wall thickness of parts,this thesis adopted the minimum distance search method.The feasibility and precision of the proposed method were verified by the measurement of multi-group CT images of precision component.This thesis systematically studied the measurement of industrial CT images of precision component.Based on the proposed WT-RSF image segmentation method,a measurement method was designed for industrial CT images of precision component and the accuracy of this method was verified by measuring several typical geometric elements.Overall,the measurement accuracy of the proposed method is better than the general software VG Studio Max,therefore,this study has engineering application value.