Point cloud reduction method based on fuzzy entropy iteration

Abstract

The invention discloses a point cloud reduction method based on fuzzy entropy iteration, which mainly aims to realize better detail features for an obtained reduced point cloud model while increasing the running efficiency of a reduction method. The method comprises the following steps of firstly, performing rapid X-Y boundary extraction on all point cloud data to keep point cloud boundary features; secondly, calculating the curvatures of all data points, grouping the data points except a boundary according to the curvatures, and calculating the quantity of data points in each group and an average curvature value; thirdly, constructing a fuzzy set of the point cloud model by using the curvatures of the data points, and calculating a minimum fuzzy entropy to obtain an optimal curvature partition threshold; and lastly, diluting the data points of which the curvatures are less than the threshold in a corresponding ratio according to different iteration times, performing iteration calculation fuzzy entropy operation on data points of which the curvatures are more than the threshold under the condition of meeting the requirement of the quantity of residual points, or retaining all data points when the requirement on quantity is not met. Through point cloud reduction, the detail features of the point cloud can be kept approximate to a point cloud prototype, and high operation efficiency is achieved.

Description

technical field [0001] The invention belongs to the technical field of three-dimensional information reconstruction, in particular to a method for using a point cloud simplification method for three-dimensional measurement of objects. Background technique [0002] Reverse engineering is a reproduction process of product design technology. The data information of the original physical model is obtained through three-dimensional measurement technology. After analysis, it can be used for industrial production or further mathematical analysis. With the improvement of the economic level, 3D machine vision has begun to enter people's field of vision, and it has become a research hotspot to study the key technologies in reverse engineering and develop a visualized three-dimensional measurement system. Today's popular 3D optical scanning technology can quickly obtain geometric data of complex surfaces, but the amount of point cloud data obtained is quite large, and storage, reconstr...

AI Technical Summary

Problems solved by technology

[0004] Purpose of the invention: Aiming at the problem that it is difficult to retain the detailed features of the point cloud while approaching the point cloud as much as possible in the process of point cloud reduction, the purpose of the present invention is to provide a point cloud reduction method based on fuzzy entropy iteration, On the premise of not affecting the subsequent 3D reconstruction effect, solve the problem of retaining incomplete details of the point cloud

Images