Research On Denoising Algorithms For 3-D Laser Point Cloud Data Of Ring Forging

Ring forgings are the basic components of all kinds of major equipment.They are widely used in aeronautics and astronautics.The key dimension parameters of the ring forgings are obtained by measuring the dimensions of the ring forgings.According to this parameter,we can measure whether the quality of forgings meets the standard or not.Therefore,the dimension measurement is an important part of the rolling process of the ring forgings.Ring forgings are affected by strong earthquake,dust,self-luminescence and heating during production.When 3-Dimensional(3D)measurements of large forgings are obtained by existing measuring methods,the measured data often contains a large amount of noise data.Therefore,in order to improve the three-dimensional measurement accuracy of ring forgings and maintain the structural characteristics of ring forgings,It is important to study ways to eliminate point cloud noise data.In this paper,the measurement of the sizes for the ring forging of 3D laser scanning and the method of data processing are studied.The specific research contents are as follows:Firstly,in order to divide the chaotic point cloud data,the point cloud data is segmented.According to the conditions of curvature constraint,the intersection point and boundary point of the two surfaces are extracted from the triangular surface formed by the sampling point and the points in its neighborhood.The other point cloud data are searched by K-D tree method.In the point cloud data of after segmentation,the data points are approximated by fitting the surface three times.Based on the curvature constraint,the noise point is regressed.A 3D point cloud segmentation and de-noising method based on curvature constraint is proposed.The smoothing and de-noising effect of the point cloud data for the ring forgings is achieved.Secondly,the selection of the neighborhood will affect the curvature of the surface.Some surface features remain inconspicuous or unreasonable.Therefore,the 3D point cloud de-noising method based on topological transformation is proposed in this paper.A topological de-noising method of point cloud data for ring forgings is studied according to topological theory.By using topological theory,the original point cloud data is transformed into 2-Dimensional(2D)space by using topological mapping function.The topological mapping function is combined with the distance weights of sampling points.The noise point is separated from the non-noise point.On this basis,a 3D point cloud segmentation and de-noising method based on topological transformation is proposed.The de-noising effect of point cloud data of ring forgings is realized.The features of ring forgings can be maintained.Finally,an experimental system for measuring the dimensions of ring forgings is established in the laboratory.The ring forging is taken as the experimental object.The point cloud data is obtained by scanning the surface of the ring forging with a 3D laser scanner.According to the experimental results,the de-noising of the point cloud data for the ring forgings can be realized by the 3D point cloud data de-noising algorithm proposed in this study.