Point cloud segmentation method and device and computer storage medium

A point cloud, to-be-segmented technology, applied in computing, image analysis, image enhancement and other directions, can solve the problems of quantization error, increased calculation amount, noise and other interference robustness, etc., to reduce the amount of calculation and calculation. The effect of small and good feature learning

Active Publication Date: 2020-02-25
PEKING UNIV
View PDF4 Cites 10 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The point cloud segmentation of the model-driven method includes steps such as edge-based segmentation, region expansion, and model matching. These steps are based on the prior information of the model and are not robust to noise and other disturbances; Point cloud segmentation, which is mainly about learning semantics from data, such as deep learning, while typical deep learning architectures require regular input data formats, such as images on a regular 2D grid or voxels on a 3D grid, in order to perform volume Therefore, for irregular 3D point clouds, they need to be converted into regular 3D voxel grids before feeding them into a typical convolutional neural network (CNN for short). or image collections, however, this will introduce quantization errors in the conversion process and lead to an excessively large amount of data and increased computation

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Point cloud segmentation method and device and computer storage medium
  • Point cloud segmentation method and device and computer storage medium
  • Point cloud segmentation method and device and computer storage medium

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0080] figure 1 A schematic flow chart of the point cloud segmentation method provided by Embodiment 1 of the present invention, as figure 1 As shown, the method includes:

[0081] S101. Obtain point cloud data to be segmented.

[0082] S102. Segment the point cloud data to be segmented by using an optimized graph convolutional neural network model.

[0083] S103. Output the segmentation result of the point cloud data to be segmented.

[0084] In practical applications, the execution subject of this embodiment may be a point cloud segmentation device. In practical applications, the point cloud segmentation device can be realized by a virtual device, such as software code, or by a physical device written with relevant execution codes, such as a U disk, or it can also be realized by integrating relevant execution codes. The physical device implementation of the code, for example, a chip, an intelligent robot, etc.

[0085] Example in combination with actual scenarios: take ...

Embodiment 2

[0088] On the basis of the first embodiment above, in the point cloud segmentation method provided by the second embodiment of the present application, the optimized graph convolutional neural network model may include multiple optimized graph convolutional feature learning layers and segmentation layers, and each The algorithm of the optimized graph convolution feature learning layer is the same;

[0089] Specifically, such as figure 2 As shown, the optimized graph convolutional neural network model can include 3 optimized graph convolution feature learning layers (the first to third graph convolution feature learning layers), and the algorithms of the 3 graph convolution feature learning layers are the same , the network model also includes a segmentation layer. In practical applications, it may also include a classification layer, which is not limited in this application.

[0090] corresponding, image 3 A schematic flow chart of the point cloud segmentation method provi...

Embodiment 3

[0130] Figure 5 It is a schematic structural diagram of the point cloud segmentation device provided in Embodiment 3 of the present application. Such as Figure 5 As shown, the device includes:

[0131] An acquisition module 610, configured to acquire point cloud data to be segmented.

[0132] The segmentation module 620 is configured to segment the point cloud data to be segmented using an optimized graph convolutional neural network model.

[0133] An output module 630, configured to output the segmentation result of the point cloud data to be segmented.

[0134] For details of the above modules, see the above figure 1 The description in the corresponding examples.

[0135] The point cloud segmentation device provided in the embodiment of the present application obtains the point cloud data to be segmented, and directly inputs the optimized graph convolutional neural network model for segmentation. Since the graph convolution operation is used, the computation amount o...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention provides a point cloud segmentation method and device and a computer storage medium. The method comprises steps of obtaining point cloud data to be segmented; segmenting the point clouddata to be segmented by adopting an optimized graph convolutional neural network model; and outputting a segmentation result of the to-be-segmented point cloud data. According to the scheme, after theto-be-segmented point cloud data is obtained, the optimized graph convolutional neural network model is directly input for segmentation; due to the fact that the graph convolution operation is adopted, the calculation amount of graph convolution is small, the calculation amount can be reduced, the optimized graph convolution neural network model can better conduct feature learning on the point cloud data, the accuracy of point cloud segmentation is improved, and the accuracy of artificial intelligence recognition is improved.

Description

technical field [0001] The present invention relates to the technical field of point cloud segmentation, in particular to a point cloud segmentation method, device and computer storage medium. Background technique [0002] Point cloud has powerful performance in three-dimensional object representation, and is an important part in applications such as unmanned driving, depth perception and semantic segmentation, and point cloud segmentation plays a very important role in it. However, the previous deep learning methods cannot learn the characteristics of point clouds very well, because the inherent disorder and irregularity of point clouds themselves greatly affect the direct end-to-end learning. [0003] Traditional point cloud segmentation methods can be divided into model-driven and data-driven. The point cloud segmentation of the model-driven method includes steps such as edge-based segmentation, region expansion, and model matching. These steps are based on the prior inf...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
Patent Type & Authority Applications(China)
IPC IPC(8): G06T7/11G06T17/20
CPCG06T7/11G06T17/20G06T2207/10012
Inventor 胡玮特古斯郭宗明
Owner PEKING UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap