VSLAM Optimization Method in Dynamic Scenes Based on YOLO-Fastest
Simultaneous localization and mapping (SLAM) is one of the core technologies for intelligent mobile robots. However, when robots perform VSLAM in dynamic scenes, dynamic objects can reduce the accuracy of mapping and localization. If deep learning-based semantic information is introduced into the SL...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-08-01
|
| Series: | Electronics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2079-9292/12/17/3538 |
| _version_ | 1797582663766245376 |
|---|---|
| author | Zijing Song Weihua Su Haiyong Chen Mianshi Feng Jiahe Peng Aifang Zhang |
| author_facet | Zijing Song Weihua Su Haiyong Chen Mianshi Feng Jiahe Peng Aifang Zhang |
| author_sort | Zijing Song |
| collection | DOAJ |
| description | Simultaneous localization and mapping (SLAM) is one of the core technologies for intelligent mobile robots. However, when robots perform VSLAM in dynamic scenes, dynamic objects can reduce the accuracy of mapping and localization. If deep learning-based semantic information is introduced into the SLAM system to eliminate the influence of dynamic objects, it will require high computing costs. To address this issue, this paper proposes a method called YF-SLAM, which is based on a lightweight object detection network called YOLO-Fastest and tightly coupled with depth geometry to remove dynamic feature points. This method can quickly identify the dynamic target area in a dynamic scene and then use depth geometry constraints to filter out dynamic feature points, thereby optimizing the VSLAM positioning performance while ensuring real-time and efficient operation of the system. This paper evaluates the proposed method on the publicly available TUM dataset and a self-made indoor dataset. Compared with ORB-SLAM2, the root-mean-square error of the Absolute Trajectory Error (ATE) can be reduced by 98.27%. The system successfully locates and constructs an accurate environmental map in a real indoor dynamic environment using a mobile robot. It is a VSLAM system that can run in real-time on low-power embedded platforms. |
| first_indexed | 2024-03-10T23:25:41Z |
| format | Article |
| id | doaj.art-5c11b10e48c54374afada6a906003811 |
| institution | Directory Open Access Journal |
| issn | 2079-9292 |
| language | English |
| last_indexed | 2024-03-10T23:25:41Z |
| publishDate | 2023-08-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Electronics |
| spelling | doaj.art-5c11b10e48c54374afada6a9060038112023-11-19T08:00:37ZengMDPI AGElectronics2079-92922023-08-011217353810.3390/electronics12173538VSLAM Optimization Method in Dynamic Scenes Based on YOLO-FastestZijing Song0Weihua Su1Haiyong Chen2Mianshi Feng3Jiahe Peng4Aifang Zhang5School of Artificial Intelligence and Data Science, Hebei University of Technology, Tianjin 300401, ChinaSchool of Mechanical Engineering, Hebei University of Technology, Tianjin 300401, ChinaSchool of Artificial Intelligence and Data Science, Hebei University of Technology, Tianjin 300401, ChinaSchool of Artificial Intelligence and Data Science, Hebei University of Technology, Tianjin 300401, ChinaSchool of Artificial Intelligence and Data Science, Hebei University of Technology, Tianjin 300401, ChinaSchool of Civil and Architectural Engineering, Beijing University of Technology, Beijing 100124, ChinaSimultaneous localization and mapping (SLAM) is one of the core technologies for intelligent mobile robots. However, when robots perform VSLAM in dynamic scenes, dynamic objects can reduce the accuracy of mapping and localization. If deep learning-based semantic information is introduced into the SLAM system to eliminate the influence of dynamic objects, it will require high computing costs. To address this issue, this paper proposes a method called YF-SLAM, which is based on a lightweight object detection network called YOLO-Fastest and tightly coupled with depth geometry to remove dynamic feature points. This method can quickly identify the dynamic target area in a dynamic scene and then use depth geometry constraints to filter out dynamic feature points, thereby optimizing the VSLAM positioning performance while ensuring real-time and efficient operation of the system. This paper evaluates the proposed method on the publicly available TUM dataset and a self-made indoor dataset. Compared with ORB-SLAM2, the root-mean-square error of the Absolute Trajectory Error (ATE) can be reduced by 98.27%. The system successfully locates and constructs an accurate environmental map in a real indoor dynamic environment using a mobile robot. It is a VSLAM system that can run in real-time on low-power embedded platforms.https://www.mdpi.com/2079-9292/12/17/3538visual SLAMobject detectiondynamic environmentsreal-time performance |
| spellingShingle | Zijing Song Weihua Su Haiyong Chen Mianshi Feng Jiahe Peng Aifang Zhang VSLAM Optimization Method in Dynamic Scenes Based on YOLO-Fastest Electronics visual SLAM object detection dynamic environments real-time performance |
| title | VSLAM Optimization Method in Dynamic Scenes Based on YOLO-Fastest |
| title_full | VSLAM Optimization Method in Dynamic Scenes Based on YOLO-Fastest |
| title_fullStr | VSLAM Optimization Method in Dynamic Scenes Based on YOLO-Fastest |
| title_full_unstemmed | VSLAM Optimization Method in Dynamic Scenes Based on YOLO-Fastest |
| title_short | VSLAM Optimization Method in Dynamic Scenes Based on YOLO-Fastest |
| title_sort | vslam optimization method in dynamic scenes based on yolo fastest |
| topic | visual SLAM object detection dynamic environments real-time performance |
| url | https://www.mdpi.com/2079-9292/12/17/3538 |
| work_keys_str_mv | AT zijingsong vslamoptimizationmethodindynamicscenesbasedonyolofastest AT weihuasu vslamoptimizationmethodindynamicscenesbasedonyolofastest AT haiyongchen vslamoptimizationmethodindynamicscenesbasedonyolofastest AT mianshifeng vslamoptimizationmethodindynamicscenesbasedonyolofastest AT jiahepeng vslamoptimizationmethodindynamicscenesbasedonyolofastest AT aifangzhang vslamoptimizationmethodindynamicscenesbasedonyolofastest |