Scattered point cloud denoising method based on normal correction and position filtering two-step method

Abstract

The invention relates to a scattered point cloud denoising method based on a normal correction and position filtering two-step method. The method comprises the following steps: initially estimating apoint cloud normal; calculating a spatial Euclidean distance and a normal deviation between the sampling point and the neighborhood point, and further solving a spatial Euclidean distance kernel function and a normal deviation kernel function; carrying out normal filtering correction, and outputting the corrected point cloud normal; taking the corrected point cloud normal as the position iterationdirection of the scattered point cloud, improving a point cloud filtering mathematical model in combination with a Gaussian kernel function adaptive to filtering parameters, and controlling a smoothing factor according to the average distance between a sampling point and a neighborhood point; and controlling the feature retention factor according to the standard deviation of the projection of thedistance deviation between the sampling point and the neighborhood point on the sampling point, substituting the corrected normal, smoothing factor and feature retention factor values into the modelto obtain the distance of the sampling point moving along the corrected normal, finally performing filtering iteration, and outputting denoised point cloud data. Feature-preserving denoising of the model is realized, and detail distortion of model features is effectively avoided.

Description

technical field [0001] The invention relates to the field of denoising preprocessing of point cloud data in reverse engineering, in particular to a method for denoising scattered point clouds based on a two-step method of normal direction correction and position filtering. Background technique [0002] Aeroengine blades, as one of the core components of aeroengines, are exposed to harsh working environments such as high temperature and high pressure for a long time, which can easily cause damage. Research on blade repair technology can reduce blade scrap rate and maintenance cost, and generate huge economic benefits. Obtaining high-precision point cloud data is a critical step in the process of reconstructing a complete blade through reverse engineering. However, due to the influence of many factors such as the defects of measuring instruments, components themselves, and the measurement environment, the massive point cloud data obtained from the measurement often contain a ...

AI Technical Summary

Problems solved by technology

[0004] Aiming at the problem that the traditional scattered point cloud denoising model easily causes model over-smoothness and local distortion, the present invention proposes a method for denoising the scattered point cloud based on the two-step method of normal correction and position filtering. The traditional scattered point cloud filtering model is improved in two aspects, and the Gaussian kernel function of constructing the adaptive value of the filtering parameters, realizes the feature-preserving denoising of the model, and effectively avoids the distortion of the model feature details

Images