Precise Measurement Of Profile And Thickness Based On Color Spectral Confocal

With the development of high-end manufacturing,it creates great requirements of the precision measurement technology.Measuring measurement technology closely related to machinery and advanced manufacturing,as the foundation of precision measurements,measurement technology not only with high accuracy,but also adaptive to different environments and materials.Real-time measurement,non-contact measurement,simultaneous measurement of multi-surface is also needed.The traditional technology of measurement,such as: capacitive micrometer,inductive micrometer,grating line displacement transducer,laser displacement sensor.However,the above measurement methods are lack of utility and adaptability of the material which hold back the further development of these measurement methods.But the color spectrum confocal sensor can meet the requirements which the above measurement technologies are unable to achieve,and realize accurate measurement of the outline of transparent and non-transparent materials.It mainly includes the following contents:1.Analyze the principle of distance measurement of the color spectrum confocal sensor and introduce the principle of thickness measurement of the transparent material and the maximum range of measurement.And the influence of the axial resolution of the sensor on the measurement accuracy and the influence of the inclination angle of the measured surface on the measurement result are analyzed theoretically.2.The traditional data filtering algorithms are difficult to remove the noise from the contour measurement data which acquire by spectral confocal sensor effectively.In order to achieve accurate measurement of the profile and thickness of both transparent and non-transparent materials.In this paper,according to the characteristics of spectral confocal sensor,we proposed an adaptive bilateral filtering algorithm based on the measurement data's intensity value.The experiments show that this algorithm not only overcomes the shortcomings of traditional bilateral filter but also preserves the edge characteristics of the measurement data,which ensures the reliability of the measurement results.3.The KD-Tree data structure is built on the bilateral filtered data.Base on this structure,MLS smooth resampling and hole filling methods is applied.Then reconstruct the three-dimensional contour of the measured object using the smoothed data.4.By measuring a standard gauge block,the measurement accuracy of the experimental device and the effectiveness of the contour measurement for the opaque material.Then thickness and profile of the transparent material are measured.According to the proposed bilateral filtering algorithm,the measurement data is filtered.After filtering,the data is meshed to reconstruct the profile of the transparent material.5.At the last,based on the working contents of this article,we acquired the profiles data of transparent material and a tungsten block by using a color spectrum confocal sensor.Then processing the measurement data by using the method which proposed in this paper.The experimental results demonstrate the effectiveness of proposed method.