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

A visual slam loop closure detection method based on distance metric learning

A distance measurement and detection method technology, applied in biological neural network models, neural architectures, navigation computing tools, etc., can solve the problems of a large amount of computation and high image feature dimensions, and achieve the effect of reducing the amount of computation

Active Publication Date: 2022-01-25
SUZHOU RUIJIU INTELLIGENT TECH CO LTD
View PDF9 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the feature dimension of the picture extracted by the CNN model is very high, so the similarity measurement requires a large amount of calculation, which is a big challenge to real-time performance.

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
  • A visual slam loop closure detection method based on distance metric learning
  • A visual slam loop closure detection method based on distance metric learning
  • A visual slam loop closure detection method based on distance metric learning

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0049]Embodiments of the present invention are described in detail below, examples of which are shown in the drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by referring to the figures are exemplary only for explaining the present invention and should not be construed as limiting the present invention.

[0050] Below in conjunction with accompanying drawing, technical scheme of the present invention is described in further detail:

[0051] The system flowchart of the method of the present invention is as figure 1 As shown, a visual SLAM loop detection method based on distance metric learning disclosed by the present invention, its specific steps are as follows:

[0052] Step 1. On the basis of the pre-trained CNN model, use the training set to optimize, and then extract image features. The type of training set and the structure of CNN are ...

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 visual SLAM loop detection method based on distance metric learning, comprising: pre-training the CNN model, using a training set for optimization; dividing pictures in the training set into k groups on average, and simultaneously inputting k CNN pre-training models , the k CNN pre-training models share parameters; use the multigroup construction method to construct multigroups for training; after constructing the multigroups of all scenes, optimize the CNN model according to the distance relationship between the multigroups; when step 4 does not construct When a suitable multiple group is obtained, the training is directly ended and the test phase is entered; the optimized CNN model is used for the actual robot loop detection application. The invention solves the technical problem of being robust to appearance changes and viewing angle changes at the same time, and reduces the calculation amount of similarity measurement.

Description

technical field [0001] The invention discloses a visual SLAM loop detection method based on distance metric learning, and relates to the technical field of robot mobile positioning. Background technique [0002] Visual SLAM (Simultaneous Localization and Mapping) is a key technology in the field of mobile robotics. In a SLAM system, a robot will build a model of its surroundings and simultaneously estimate its trajectory. A typical visual SLAM system usually consists of the following modules: visual odometry, back-end optimization, loop detection and mapping. Among them, the loop detection module is used to automatically detect whether the robot has ever been to a certain place. If the loop closure is successfully detected, the robot will be able to provide additional constraints for the back-end optimization and reduce the optimization error. [0003] During actual navigation, the environment around the robot may change. These changes can be divided into changes in appe...

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): G01C21/20G06N3/04
CPCG01C21/20G06N3/045
Inventor 高瑜陈良
Owner SUZHOU RUIJIU INTELLIGENT TECH CO LTD