3D Semantic VSLAM of Indoor Environment Based on Mask Scoring RCNN
In view of existing Visual SLAM (VSLAM) algorithms when constructing semantic map of indoor environment, there are problems with low accuracy and low label classification accuracy when feature points are sparse. This paper proposed a 3D semantic VSLAM algorithm called BMASK-RCNN based on Mask Scorin...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2020-01-01
|
| Series: | Discrete Dynamics in Nature and Society |
| Online Access: | http://dx.doi.org/10.1155/2020/5916205 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832560044258885632 |
|---|---|
| author | Chongben Tao Yufeng Jin Feng Cao Zufeng Zhang Chunguang Li Hanwen Gao |
| author_facet | Chongben Tao Yufeng Jin Feng Cao Zufeng Zhang Chunguang Li Hanwen Gao |
| author_sort | Chongben Tao |
| collection | DOAJ |
| description | In view of existing Visual SLAM (VSLAM) algorithms when constructing semantic map of indoor environment, there are problems with low accuracy and low label classification accuracy when feature points are sparse. This paper proposed a 3D semantic VSLAM algorithm called BMASK-RCNN based on Mask Scoring RCNN. Firstly, feature points of images are extracted by Binary Robust Invariant Scalable Keypoints (BRISK) algorithm. Secondly, map points of reference key frame are projected to current frame for feature matching and pose estimation, and an inverse depth filter is used to estimate scene depth of created key frame to obtain camera pose changes. In order to achieve object detection and semantic segmentation for both static objects and dynamic objects in indoor environments and then construct dense 3D semantic map with VSLAM algorithm, a Mask Scoring RCNN is used to adjust its structure partially, where a TUM RGB-D SLAM dataset for transfer learning is employed. Semantic information of independent targets in scenes provides semantic information including categories, which not only provides high accuracy of localization but also realizes the probability update of semantic estimation by marking movable objects, thereby reducing the impact of moving objects on real-time mapping. Through simulation and actual experimental comparison with other three algorithms, results show the proposed algorithm has better robustness, and semantic information used in 3D semantic mapping can be accurately obtained. |
| format | Article |
| id | doaj-art-7f28c5b77d0c495bb90a03888a33691e |
| institution | Kabale University |
| issn | 1026-0226 1607-887X |
| language | English |
| publishDate | 2020-01-01 |
| publisher | Wiley |
| record_format | Article |
| series | Discrete Dynamics in Nature and Society |
| spelling | doaj-art-7f28c5b77d0c495bb90a03888a33691e2025-02-03T01:28:29ZengWileyDiscrete Dynamics in Nature and Society1026-02261607-887X2020-01-01202010.1155/2020/591620559162053D Semantic VSLAM of Indoor Environment Based on Mask Scoring RCNNChongben Tao0Yufeng Jin1Feng Cao2Zufeng Zhang3Chunguang Li4Hanwen Gao5School of Electronic and Information Engineering, Suzhou University of Science and Technology, Suzhou 215009, Jiangsu, ChinaSchool of Electronic and Information Engineering, Suzhou University of Science and Technology, Suzhou 215009, Jiangsu, ChinaSchool of Computer and Information Technology, Shanxi University, Taiyuan 030006, Shanxi, ChinaDepartment of Automation, Tsinghua University, Beijing 100084, ChinaSchool of Computer Information and Engineering, Changzhou Institute of Technology, Changzhou 213002, Jiangsu, ChinaSchool of Electronic and Information Engineering, Suzhou University of Science and Technology, Suzhou 215009, Jiangsu, ChinaIn view of existing Visual SLAM (VSLAM) algorithms when constructing semantic map of indoor environment, there are problems with low accuracy and low label classification accuracy when feature points are sparse. This paper proposed a 3D semantic VSLAM algorithm called BMASK-RCNN based on Mask Scoring RCNN. Firstly, feature points of images are extracted by Binary Robust Invariant Scalable Keypoints (BRISK) algorithm. Secondly, map points of reference key frame are projected to current frame for feature matching and pose estimation, and an inverse depth filter is used to estimate scene depth of created key frame to obtain camera pose changes. In order to achieve object detection and semantic segmentation for both static objects and dynamic objects in indoor environments and then construct dense 3D semantic map with VSLAM algorithm, a Mask Scoring RCNN is used to adjust its structure partially, where a TUM RGB-D SLAM dataset for transfer learning is employed. Semantic information of independent targets in scenes provides semantic information including categories, which not only provides high accuracy of localization but also realizes the probability update of semantic estimation by marking movable objects, thereby reducing the impact of moving objects on real-time mapping. Through simulation and actual experimental comparison with other three algorithms, results show the proposed algorithm has better robustness, and semantic information used in 3D semantic mapping can be accurately obtained.http://dx.doi.org/10.1155/2020/5916205 |
| spellingShingle | Chongben Tao Yufeng Jin Feng Cao Zufeng Zhang Chunguang Li Hanwen Gao 3D Semantic VSLAM of Indoor Environment Based on Mask Scoring RCNN Discrete Dynamics in Nature and Society |
| title | 3D Semantic VSLAM of Indoor Environment Based on Mask Scoring RCNN |
| title_full | 3D Semantic VSLAM of Indoor Environment Based on Mask Scoring RCNN |
| title_fullStr | 3D Semantic VSLAM of Indoor Environment Based on Mask Scoring RCNN |
| title_full_unstemmed | 3D Semantic VSLAM of Indoor Environment Based on Mask Scoring RCNN |
| title_short | 3D Semantic VSLAM of Indoor Environment Based on Mask Scoring RCNN |
| title_sort | 3d semantic vslam of indoor environment based on mask scoring rcnn |
| url | http://dx.doi.org/10.1155/2020/5916205 |
| work_keys_str_mv | AT chongbentao 3dsemanticvslamofindoorenvironmentbasedonmaskscoringrcnn AT yufengjin 3dsemanticvslamofindoorenvironmentbasedonmaskscoringrcnn AT fengcao 3dsemanticvslamofindoorenvironmentbasedonmaskscoringrcnn AT zufengzhang 3dsemanticvslamofindoorenvironmentbasedonmaskscoringrcnn AT chunguangli 3dsemanticvslamofindoorenvironmentbasedonmaskscoringrcnn AT hanwengao 3dsemanticvslamofindoorenvironmentbasedonmaskscoringrcnn |