Data processing method, apparatus, system and storage media

Patent metadata

Field	Value
Publication number	US-10706567-B2
Application number	US-201816165888-A
Country	US
Kind code	B2
Filing date	Oct 19, 2018
Priority date	Oct 20, 2017
Publication date	Jul 7, 2020
Grant date	Jul 7, 2020

Abstract

A data processing method includes determining feature points in first point cloud data and feature points in second point cloud data, the first point cloud data and the second point cloud data being used for representing different parts of a same object; performing feature matching between the first point cloud data and the second point cloud data to determine feature points satisfying feature matching condition(s) between the first point cloud data and the second point cloud data, and form a plurality of feature point pairs; determining a transformation matrix in which spatial distances between feature points in one or more feature point pairs of the plurality of feature point pairs conform to a proximity condition; and performing coordinate transformation on the one or more feature point pairs using the transformation matrix to register the first point cloud data with the second point cloud data.

First claim

What is claimed is:

1. A method implemented by one or more computing devices, the method comprising:
determining feature points in first point cloud data and feature points in second point cloud data, the first point cloud data and the second point cloud data being used for representing different parts of a same object;
performing feature matching between the first point cloud data and the second point cloud data to determine feature points that satisfy a feature matching condition between the first point cloud data and the second point cloud data, and form a plurality of feature point pairs;
for one or more feature point pairs of the plurality of feature point pairs, determining a transformation matrix in which a spatial distance between feature points in the feature point pairs conform to a proximity condition;
performing coordinate transformation on the one or more feature point pairs of the plurality of feature point pairs using the transformation matrix to register the first point cloud data with the second point cloud data to obtain registered first point cloud data and registered second point cloud data; and
using the registered first point cloud data and the registered second point cloud data as the first point cloud data to re-determine the feature points in the first point cloud data until a three-dimensional point cloud data model of the object is obtained.

2. The method of claim 1 , wherein determining the feature points in the first point cloud data and the feature points in the second point cloud data comprises:
extracting a feature point that meets a selection criterion of a specified feature from the first point cloud data according to the selection criterion; and
extracting a feature point matching the selection criterion from the second point cloud data.

3. The method of claim 2 , wherein the specified feature includes at least a geometric feature or a color feature.

4. The method of claim 1 , wherein performing the feature matching between the first point cloud data and the second point cloud data to determine the feature points that satisfy the feature matching condition between the first point cloud data and the second point cloud data, and form the plurality of feature point pairs comprises:
obtaining a first feature point in the first point cloud data, and using a feature value of the first feature point to find a second feature point in the second point cloud data, a feature of the second feature point and the feature value of the first feature point satisfying a feature value threshold condition;
searching for a third feature point in the first point cloud data using the feature value of the second feature point, a feature value of the third feature point and the feature value of the second feature point satisfying the feature value threshold condition; and
determining the first feature point and the second feature point as feature points that meet the feature matching condition when the first feature point and the third feature point coincide, and form the plurality of feature point pairs.

5. The method of claim 1 , wherein: for the one or more feature point pairs of the plurality of feature point pairs, determining the transformation matrix in which the spatial distance between the feature points in the feature point pairs conform to the proximity condition comprises:
constructing an evaluation model of the one or more feature point pairs of the plurality of feature point pairs based on a spatial distance and a precision control parameter;
iteratively processing the spatial distance using the evaluation model under control of the precision control parameter to obtain effective feature point pairs in the feature point pairs and a transformation matrix among the effective feature point pairs, and treating transformation matrix between the effective feature point pairs as a transformation matrix in which the spatial distance between feature points in the feature point pairs conforms to the proximity condition.

6. The method of claim 5 , wherein iteratively processing the spatial distance using the evaluation model under control of the precision control parameter to obtain the effective feature point pairs in the feature point pairs and the transformation matrix among the effective feature point pairs comprises:
reducing a value of the precision control parameter, and constructing a new evaluation model using the evaluation model and the reduced precision control parameter; and
finding a solution for the spatial distance of the feature point pairs using the new evaluation model, and continuing to reduce the precision control parameter for iterative processing until the value of the spatial distance takes a minimum, to obtain the effective feature point pairs and the transformation matrix between the effective feature point pairs.

7. An apparatus comprising:
one or more processors;
memory;
a feature acquisition module stored in the memory and executable by the one or more processors to determine feature points in first point cloud data and feature points in second point cloud data, the first point cloud data and the second point cloud data being used for representing different parts of a same object;
a feature matching module stored in the memory and executable by the one or more processors to perform feature matching between the first point cloud data and the second point cloud data to determine feature points that satisfy a feature matching condition between the first point cloud data and the second point cloud data, and form a plurality of feature point pairs;
a feature point pair selection module stored in the memory and executable by the one or more processors to determine a transformation matrix in which spatial distances between feature points in one or more feature point pairs of the plurality of feature point pairs conform to a proximity condition for the feature point pairs;
a data registration module stored in the memory and executable by the one or more processors to perform coordinate transformation on the one or more feature point pairs of the plurality of feature point pairs using the transformation matrix to register the first point cloud data with the second point cloud data to obtain registered first point cloud data and registered second point cloud data; and
an iterative registration module configured to use the registered first point cloud data and the registered second point cloud data as the first point cloud data to re-determine the feature points in the first point cloud data until a three-dimensional point cloud data model of the object is obtained.

8. The apparatus of claim 7 , wherein the feature acquisition module comprises:
a first feature point selection unit configured to extract a feature point that meets a selection criterion of a specified feature from the first point cloud data according to the selection criterion; and
a second feature point selection unit configured to extract a feature point matching the selection criterion from the second point cloud data.

9. The apparatus of claim 8 , apparatus wherein the specified feature comprises at least a geometric feature or a color feature.

10. The apparatus of claim 7 , wherein the feature matching module comprises:
a first feature point searching unit configured to obtain a first feature point in the first point cloud data, and use a feature value of the first feature point to find a second feature point in the second point cloud data, a feature of the second feature point and the feature value of the first feature point satisfying a feature value threshold condition;
a second feature point searching unit configured to search for a third feature point in the first point cloud data using the

Assignees

• Alibaba Group Holding Ltd

Inventors

• Wang Bin
• Yu Jingming
• Feng Xiaoduan
• Pan Pan
• Jin Rong

Classifications

• G06V10/44

  Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components · CPC title

• G06T19/00 Primary

  Manipulating three-dimensional [3D] models or images for computer graphics · CPC title

• G06T7/33 Primary

  using feature-based methods · CPC title

• G06V10/757

  Matching configurations of points or features · CPC title

• G06T2207/20164

  Salient point detection; Corner detection · CPC title