Method for extracting circles from complex object point clouds and scanning device

Abstract

The invention discloses a method for extracting circles from complex object point clouds. The method comprises the following steps that: complex object point clouds are visualized, the data points ofa slice reference plane are manually selected, and the feature vector of the reference slice plane is obtained through a plane fitting method; the complex object point clouds are sliced by using two slicing planes with the interval of delta, so that three-dimensional tangent plane point clouds can be obtained, and the two slicing planes are parallel to the reference slicing plane; the three-dimensional tangent plane point clouds are converted into two-dimensional tangent plane point clouds by adopting a projection method; and finally, circles are extracted according to the two-dimensional tangent plane point clouds, and circle parameters are calculated. According to the method, data in a certain range are integrated into a small range, so that the problem of processing the three-dimensional data is converted into the problem of processing the two-dimensional data; and therefore, a data search range is reduced, calculation is simplified, and efficiency is improved.

Description

technical field [0001] The invention relates to the technical field of three-dimensional scanning, in particular to a method for extracting circles from point clouds of complex objects and a scanning device. Background technique [0002] The existing industrial size detection methods generally use CCD / CMOS image sensors for image acquisition, and using this method needs to consider many factors, such as light source, lens selection, sensor selection, installation layout, automation integration, environmental considerations, workpiece state changes and so on. Any deviation in any aspect will affect the imaging quality, thereby affecting the imaging results. Moreover, industrial parts are three-dimensional parts. In view of the difficulties in parameter detection of deep-hole parts, using binocular cameras for three-dimensional imaging and binocular vision three-dimensional reconstruction is one of the important research contents in the field of computer vision. In precision ...

AI Technical Summary

Problems solved by technology

[0004] During the scanning process of point cloud data, for larger objects, the scanning equipment cannot scan the whole picture of the object at one time, resulting in scattered 3D data

Images