Consistent point shift registration method based on high-dimensional representation

A base point and registration technology, applied in image analysis, instrumentation, calculation, etc.

Active Publication Date: 2020-04-03
XI'AN INST OF OPTICS & FINE MECHANICS - CHINESE ACAD OF SCI +1
View PDF3 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0017] In order to solve the problem of the existing consistent point drift registration algorithm in the background technology, the present invention proposes a consistent point drift registration method based on high-dimensional expression, which can realize point cloud under the distortion of rotation, deformation and noise. Highly accurate and fast registration

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
  • Consistent point shift registration method based on high-dimensional representation
  • Consistent point shift registration method based on high-dimensional representation
  • Consistent point shift registration method based on high-dimensional representation

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0090]In this embodiment, the fish model data set is used for experiments. The hardware configuration is 3.4GHz Inter Core i3 CPU, 4GB RAM, and the experimental platform is MATLAB (R2014a).

[0091] 1) Normalize the two point clouds collected from the fish model dataset respectively; the two point clouds are: data point cloud X N×D =(x 1 ,...,x N ) and model point cloud Y M×D =(y 1 ,...,y M );

[0092] Among them, D represents the point cloud dimension, N represents the number of data point clouds; M represents the number of model point clouds;

[0093] 1.1) Solve the mean value of the data point cloud separately and standard deviation σ x and the mean of the model point cloud and standard deviation σ y ;

[0094] 1.2) Subtract the mean value from the coordinates of the data point cloud and the model point cloud respectively and divide by the standard deviation. The processed data point cloud is: The model point cloud is

[0095] 2) Select two base point sets...

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 discloses a consistent point drift registration method based on high-dimensional expression. The method comprises the following steps: to begin with, normalizing collected point cloud; then, selecting a base point set by utilizing a local feature descriptor; calculating relative position relation between points and the base point set; fusing a relative structure of the point into anoriginal coordinate of the point, and carrying out dimension rising on the point cloud; and finally, carrying out high-dimensional point cloud registration modeling into a maximum likelihood estimation problem, converting fuzzy correspondence relation between points into calculation of posterior probability, converting the non-rigid transformation into calculation of a velocity field weight coefficient, and carrying out smooth constraint on a velocity field through Tikhonov regularization. Experiment results show that high-precision and fast point cloud registration can be realized under conditions of degeneration of rotation, deformation and noise; and the achievements can be applied to research and application in related fields of virtual reality and human posture tracking and the like.

Description

technical field [0001] The invention relates to a consistent point drift registration method based on high-dimensional expression, which is applied to three-dimensional reconstruction, virtual reality and human body posture estimation. Background technique [0002] Point cloud registration is one of the basic research topics in the field of computer vision. In particular, with the advent of more and more cheap depth detectors now, point cloud registration has received increasing attention. For example, in the fields of virtual reality, stereo vision matching, and artificial intelligence, it is often necessary to fuse point clouds (or images) obtained from different angles or at different times to obtain a complete scene. Due to the influence of perspective changes, scene deformation and noise in general scene fusion, point cloud matching becomes complicated and time-consuming, and it is difficult to have a perfect solution. Point cloud registration has always been a challe...

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 Patents(China)
IPC IPC(8): G06T7/33G06T7/37
CPCG06T7/344G06T7/37
Inventor 周祚峰黄会敏曹剑中王亚楠
Owner XI'AN INST OF OPTICS & FINE MECHANICS - CHINESE ACAD OF SCI
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