Unlock instant, AI-driven research and patent intelligence for your innovation.

Three-dimensional point cloud affine registration method and system based on pseudo Huber loss function

A technology of 3D point cloud and loss function, which is applied in image data processing, instrumentation, 3D modeling, etc. It can solve problems such as wrong registration results, affect the accuracy of point cloud registration, and poor registration results, so as to improve the accuracy Effect

Active Publication Date: 2020-11-17
NORTHWEST UNIV
View PDF7 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] At present, most of the existing point cloud affine registration algorithms have poor registration results and application limitations, such as poor registration accuracy for data missing or containing a large number of noise points and outliers
Due to the physical limitation of the point cloud data acquisition sensor and the influence of noise and other factors, the actual point cloud collected has data missing, noise points and outliers, which will affect the registration accuracy of the point cloud and even obtain wrong registration results.

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
  • Three-dimensional point cloud affine registration method and system based on pseudo Huber loss function
  • Three-dimensional point cloud affine registration method and system based on pseudo Huber loss function
  • Three-dimensional point cloud affine registration method and system based on pseudo Huber loss function

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0046] The 3D point cloud registration technology is to match the collected point cloud data, find a series of spatial transformation relations of two sets of point cloud datasets, and through these spatial transformations, make the corresponding points of the two point cloud datasets in the spatial position As consistent as possible, it is mainly divided into rigid registration and non-rigid registration. Among them, the simplest non-rigid registration is affine registration.

[0047] The purpose of 3D point cloud affine registration is to establish the spatial correspondence between two point clouds and find the optimal affine transformation between them. Usually, 3D point cloud affine registration includes solving two contents: one is to establish the correspondence between two point clouds; the other is to solve the affine transformation between two point clouds.

[0048] given shape point cloud and model point cloud where N p and N q Both are positive integers, assu...

Embodiment 2

[0111] On the basis of the first embodiment above, this embodiment provides a 3D point cloud affine registration system based on pseudo-Huber loss function, please refer to image 3 , image 3 It is a schematic structural diagram of another 3D point cloud affine registration system based on a pseudo-Huber loss function provided by an embodiment of the present invention. The system includes:

[0112] Data acquisition module 1, used to acquire shape point cloud and model point cloud

[0113] The model building module 2 is used to establish a three-dimensional point cloud affine registration model based on a pseudo-Huber loss function according to the shape point cloud and the model point cloud; the built model is expressed as:

[0114]

[0115] Among them, A represents the affine matrix, t represents the translation vector, c represents the spatial correspondence between the shape point cloud and the model point cloud, b represents the outlier threshold, and p i Represe...

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 three-dimensional point cloud affine registration method and system based on a pseudo Huber loss function. The method comprises the following steps: obtaining a shape point cloud and a model point cloud; establishing a three-dimensional point cloud affine registration model based on a pseudo Huber loss function according to the shape point cloud and the model point cloud;and carrying out optimization solution on the three-dimensional point cloud affine registration model to obtain optimal affine transformation. According to the three-dimensional point cloud affine registration method based on the pseudo Huber loss function provided by the invention, the influence of data missing, noise points or external points on a three-dimensional point cloud affine registration result can be suppressed, and the precision of affine registration is improved.

Description

technical field [0001] The invention belongs to the technical field of three-dimensional point cloud data processing, and in particular relates to a three-dimensional point cloud affine registration method and system based on a pseudo-Huber loss function. Background technique [0002] With the rapid development of data acquisition technology and computer technology, the research object of computer vision technology has gradually changed from two-dimensional images to three-dimensional point cloud data. In practice, affine registration is required due to extensive affine transformations in 3D point cloud data. However, most of the existing affine registration methods have the problem of poor accuracy or the need for additional data information. For example, Feldmar and Ayache first perform rigid body registration on point clouds, and then introduce normal vectors and principal curvatures at each coordinate point for calculation Affine transformation, but this method has poor...

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): G06T7/30G06T17/00
CPCG06T7/30G06T17/00G06T2207/10028
Inventor 汪霖郭佳琛樊萍周延刘成姜博李艳艳陈晓璇汪照阳高歌
Owner NORTHWEST UNIV