Visual SLAM Based on Lightweight Dynamic Object Detection

被引:0
|
作者
Jiang, Changjiang [1 ]
Lin, Tong [1 ]
Tan, Li [1 ]
Zhao, Changhao [1 ]
机构
[1] Chongqing Univ Posts & Telecommun, Sch Automat, Chongqing, Peoples R China
关键词
visual synchronous positioning and mapping; lightweight object detection network; optical flow method; dynamic point culling;
D O I
10.1109/CCDC58219.2023.10327139
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
The traditional visual synchronous localization and mapping algorithm (visual SLAM) assumes that the external environment is static. The system positioning accuracy drops when it is applied in the dynamic environment. In this paper, CBAM (Convolutional Block Attention Module) and ShuffleNetV2 are fused to improve the network feature extraction ability, and the fused network is used to reconstruct the backbone network of YOLOv5s, reducing the amount of model parameters and improving the algorithm reasoning speed. The improved object detection network and optical flow method are combined to eliminate the dynamic feature points extracted from the front end of the SLAM. And the remaining static feature points are used for inter frame matching to solve the camera pose. The experimental results on TUM dataset show that this method has improved the positioning accuracy of the system in indoor dynamic environment, and has achieved some improvement in real-time and accuracy compared with other classical dynamic visul SLAM algorithms.
引用
收藏
页码:1158 / 1163
页数:6
相关论文
共 50 条
  • [41] Online Object Detection and Localization on Stereo Visual SLAM System
    Pire, Taihu
    Corti, Javier
    Grinblat, Guillermo
    JOURNAL OF INTELLIGENT & ROBOTIC SYSTEMS, 2020, 98 (02) : 377 - 386
  • [42] Online Object Detection and Localization on Stereo Visual SLAM System
    Taihú Pire
    Javier Corti
    Guillermo Grinblat
    Journal of Intelligent & Robotic Systems, 2020, 98 : 377 - 386
  • [43] A Lightweight Neural Network for Loop Closure Detection in Indoor Visual SLAM
    Deyang Zhou
    Yazhe Luo
    Qinhan Zhang
    Ying Xu
    Diansheng Chen
    Xiaochuan Zhang
    International Journal of Computational Intelligence Systems, 16
  • [44] A Lightweight Neural Network for Loop Closure Detection in Indoor Visual SLAM
    Zhou, Deyang
    Luo, Yazhe
    Zhang, Qinhan
    Xu, Ying
    Chen, Diansheng
    Zhang, Xiaochuan
    INTERNATIONAL JOURNAL OF COMPUTATIONAL INTELLIGENCE SYSTEMS, 2023, 16 (01)
  • [45] DO-SLAM: research and application of semantic SLAM system towards dynamic environments based on object detection
    Wei, Yaoguang
    Zhou, Bingqian
    Duan, Yunhong
    Liu, Jincun
    An, Dong
    APPLIED INTELLIGENCE, 2023, 53 (24) : 30009 - 30026
  • [46] DO-SLAM: research and application of semantic SLAM system towards dynamic environments based on object detection
    Yaoguang Wei
    Bingqian Zhou
    Yunhong Duan
    Jincun Liu
    Dong An
    Applied Intelligence, 2023, 53 : 30009 - 30026
  • [47] Indoor Dynamic Scene Visual SLAM Based On Human Detection And Geometric Constraints
    He, Jiantao
    He, Li
    Zhu, Haifei
    Xiang, Chaoqun
    Guan, Yisheng
    PROCEEDINGS OF THE 33RD CHINESE CONTROL AND DECISION CONFERENCE (CCDC 2021), 2021, : 1652 - 1657
  • [48] A dynamic detection and data association method based on probabilistic models for visual SLAM
    Zhang, Jianbo
    Yuan, Liang
    Ran, Teng
    Peng, Song
    Tao, Qing
    Xiao, Wendong
    Cui, Jianping
    DISPLAYS, 2024, 82
  • [49] Robust visual SLAM algorithm based on target detection and clustering in dynamic scenarios
    Gan, Fubao
    Xu, Shanyong
    Jiang, Linya
    Liu, Yuwen
    Liu, Quanzeng
    Lan, Shihao
    FRONTIERS IN NEUROROBOTICS, 2024, 18
  • [50] A Dynamic Scene Vision SLAM Method Incorporating Object Detection and Object Characterization
    Guan, Hongliang
    Qian, Chengyuan
    Wu, Tingsong
    Hu, Xiaoming
    Duan, Fuzhou
    Ye, Xinyi
    SUSTAINABILITY, 2023, 15 (04)