An Improved Visual SLAM Algorithm Based on Graph Neural Network
Feature extraction and matching are irreplaceable parts of a typical visual simultaneous localization and mapping (VSLAM) algorithm. A variety of different approaches (e.g., ORB, Superpoint, GCNv2, etc.) have been proposed for effective feature extraction and matching. However, as far as we know, su...
Main Authors: | , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
IEEE
2023-01-01
|
Series: | IEEE Access |
Subjects: | |
Online Access: | https://ieeexplore.ieee.org/document/10243031/ |
_version_ | 1797672864544980992 |
---|---|
author | Wei Wang Tao Xu Kaisheng Xing Jinhui Liu Mengyuan Chen |
author_facet | Wei Wang Tao Xu Kaisheng Xing Jinhui Liu Mengyuan Chen |
author_sort | Wei Wang |
collection | DOAJ |
description | Feature extraction and matching are irreplaceable parts of a typical visual simultaneous localization and mapping (VSLAM) algorithm. A variety of different approaches (e.g., ORB, Superpoint, GCNv2, etc.) have been proposed for effective feature extraction and matching. However, as far as we know, such methods still face great challenges in extreme angle and texture sparse scenarios. In this paper, an Improved Visual SLAM Algorithm Based on Graph Neural Network (GNNI-VSLAM), inspired by the strong robustness of the graph neural network in the field of image matching, is proposed to solve the problem of feature point extraction difficulties in sparsely textures scenes and feature point matching difficulties at extreme angles. First, the a priori location estimation feature extraction network is proposed to obtain fast and uniform detection and description of image feature points by a priori location estimation and to build accurate and real feature point information. Second, the feature matching network of the graph attention mechanism is proposed to aggregate feature point information through the neural network of the message passing graph, and then use the self and joint attention mechanism for adjacent image frame weighted feature matching. Then, the feature extraction and neural network are merged with the back-end nonlinear optimization, closed-loop correction and local mapping modules of the ORB-SLAM2 system to propose a complete monocular vision GNNI-VSLAM system. Finally, the proposed algorithm is verified by the public TUM dataset and the experimental results show that the absolute trajectory error of the proposed algorithm is reduced by 29.7 % compared to the GCNv2-SLAM algorithm, which shows a good mapping capabilities. |
first_indexed | 2024-03-11T21:36:23Z |
format | Article |
id | doaj.art-01bd66cc1c294ffe90db1d40d8bdc260 |
institution | Directory Open Access Journal |
issn | 2169-3536 |
language | English |
last_indexed | 2024-03-11T21:36:23Z |
publishDate | 2023-01-01 |
publisher | IEEE |
record_format | Article |
series | IEEE Access |
spelling | doaj.art-01bd66cc1c294ffe90db1d40d8bdc2602023-09-26T23:00:18ZengIEEEIEEE Access2169-35362023-01-011110236610238010.1109/ACCESS.2023.331271410243031An Improved Visual SLAM Algorithm Based on Graph Neural NetworkWei Wang0https://orcid.org/0009-0008-6535-6565Tao Xu1https://orcid.org/0009-0006-7709-4756Kaisheng Xing2Jinhui Liu3Mengyuan Chen4https://orcid.org/0000-0003-3264-0494School of Electrical and Electronic Engineering, Anhui Institute of Information Technology, Wuhu, ChinaSchool of Electrical Engineering, Anhui Polytechnic University, Wuhu, ChinaSchool of Electrical and Electronic Engineering, Anhui Institute of Information Technology, Wuhu, ChinaSchool of Electrical Engineering, Anhui Polytechnic University, Wuhu, ChinaAnhui Engineering University Industrial Innovation Technology Research Co., Ltd., Wuhu, ChinaFeature extraction and matching are irreplaceable parts of a typical visual simultaneous localization and mapping (VSLAM) algorithm. A variety of different approaches (e.g., ORB, Superpoint, GCNv2, etc.) have been proposed for effective feature extraction and matching. However, as far as we know, such methods still face great challenges in extreme angle and texture sparse scenarios. In this paper, an Improved Visual SLAM Algorithm Based on Graph Neural Network (GNNI-VSLAM), inspired by the strong robustness of the graph neural network in the field of image matching, is proposed to solve the problem of feature point extraction difficulties in sparsely textures scenes and feature point matching difficulties at extreme angles. First, the a priori location estimation feature extraction network is proposed to obtain fast and uniform detection and description of image feature points by a priori location estimation and to build accurate and real feature point information. Second, the feature matching network of the graph attention mechanism is proposed to aggregate feature point information through the neural network of the message passing graph, and then use the self and joint attention mechanism for adjacent image frame weighted feature matching. Then, the feature extraction and neural network are merged with the back-end nonlinear optimization, closed-loop correction and local mapping modules of the ORB-SLAM2 system to propose a complete monocular vision GNNI-VSLAM system. Finally, the proposed algorithm is verified by the public TUM dataset and the experimental results show that the absolute trajectory error of the proposed algorithm is reduced by 29.7 % compared to the GCNv2-SLAM algorithm, which shows a good mapping capabilities.https://ieeexplore.ieee.org/document/10243031/Simultaneous localization and mapping (SLAM)large viewing angle motiongraph neural networkgraph attention mechanismmessage pass neural network |
spellingShingle | Wei Wang Tao Xu Kaisheng Xing Jinhui Liu Mengyuan Chen An Improved Visual SLAM Algorithm Based on Graph Neural Network IEEE Access Simultaneous localization and mapping (SLAM) large viewing angle motion graph neural network graph attention mechanism message pass neural network |
title | An Improved Visual SLAM Algorithm Based on Graph Neural Network |
title_full | An Improved Visual SLAM Algorithm Based on Graph Neural Network |
title_fullStr | An Improved Visual SLAM Algorithm Based on Graph Neural Network |
title_full_unstemmed | An Improved Visual SLAM Algorithm Based on Graph Neural Network |
title_short | An Improved Visual SLAM Algorithm Based on Graph Neural Network |
title_sort | improved visual slam algorithm based on graph neural network |
topic | Simultaneous localization and mapping (SLAM) large viewing angle motion graph neural network graph attention mechanism message pass neural network |
url | https://ieeexplore.ieee.org/document/10243031/ |
work_keys_str_mv | AT weiwang animprovedvisualslamalgorithmbasedongraphneuralnetwork AT taoxu animprovedvisualslamalgorithmbasedongraphneuralnetwork AT kaishengxing animprovedvisualslamalgorithmbasedongraphneuralnetwork AT jinhuiliu animprovedvisualslamalgorithmbasedongraphneuralnetwork AT mengyuanchen animprovedvisualslamalgorithmbasedongraphneuralnetwork AT weiwang improvedvisualslamalgorithmbasedongraphneuralnetwork AT taoxu improvedvisualslamalgorithmbasedongraphneuralnetwork AT kaishengxing improvedvisualslamalgorithmbasedongraphneuralnetwork AT jinhuiliu improvedvisualslamalgorithmbasedongraphneuralnetwork AT mengyuanchen improvedvisualslamalgorithmbasedongraphneuralnetwork |