Related references
Note: Only part of the references are listed.
Article
Robotics
Zhanyuan Chang et al.
Summary: The paper presents a robust and real-time RGB-D SLAM algorithm based on ORBSLAM3. It addresses the challenges faced in feature point matching-based SLAM in complex and changeable indoor environments. The proposed algorithm improves pose estimation accuracy in the presence of dynamic objects and avoids tracking loss in weak texture scenes. Experimental results show that the algorithm achieves significantly improved performance compared to existing algorithms in terms of accuracy and real-time capability.
JOURNAL OF FIELD ROBOTICS
(2023)
Article
Chemistry, Analytical
Qiuyu Zang et al.
Summary: Recent developments in robotics have led to an increased demand for visual SLAM, however, dynamic objects pose a major problem reducing localization accuracy. This study proposes an adaptive feature point selection system for outdoor dynamic environments to address this issue. The system uses YOLOv5s to identify dynamic objects, selects feature points based on their presence and occupancy in the frame, and determines dynamic regions using Lucas-Kanade optical flow and RANSAC algorithm. Evaluation on the KITTI dataset shows a significant reduction in absolute and relative trajectory errors compared to other systems.
Article
Environmental Sciences
Chaofeng Yuan et al.
Summary: This study proposes a point-line SLAM system based on dynamic environments. It obtains dynamic region features through detection and segmentation of dynamic regions. The separation of dynamic and static objects is achieved through a geometric constraint method for matching line segments and a dynamic feature tracking method based on Bayesian theory, improving the robustness and accuracy of the SLAM system.
Article
Computer Science, Artificial Intelligence
Yingchun Fan et al.
Summary: The paper proposes Blitz-SLAM, a semantic SLAM system working in indoor dynamic environments, which removes noise blocks in the local point cloud by combining mask, RGB and depth images and can work robustly in dynamic environments to generate a clean and accurate global point cloud map simultaneously.
PATTERN RECOGNITION
(2022)
Article
Chemistry, Analytical
Ying Sun et al.
Summary: SLAM technology is used for locating and mapping in unknown environments, but the constructed maps often lack readability and interactivity, making it difficult to grasp the information accurately. To enable intelligent robots to interact meaningfully with their environment, it is necessary to understand both the geometric and semantic properties of the scene. The proposed method reduces absolute positional errors, constructs dense semantic point cloud maps, and segments point cloud models of objects in the environment with high accuracy.
Proceedings Paper
Computer Science, Artificial Intelligence
Yifu Zhang et al.
Summary: This method improves the performance of multi-object tracking by associating almost every detection box, effectively solving the problem of true object missing and fragmented trajectories caused by low score detection boxes being discarded. Applied to multiple trackers, this method consistently achieves improvement on IDF1 score.
COMPUTER VISION, ECCV 2022, PT XXII
(2022)
Article
Robotics
Berta Bescos et al.
Summary: The paper introduces DynaSLAM II, a visual SLAM system for stereo and RGB-D camera configurations with tight integration of multi-object tracking ability, utilizing instance semantic segmentation and ORB features to track dynamic objects. The system not only provides rich clues for scene understanding but also benefits camera tracking.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2021)
Article
Robotics
Yuzhen Liu et al.
Summary: The proposed switching-coupled back-end solution offers a flexible approach to address the limitations in SLOT problem. By considering the uncertainty, observation quality, and prior information of objects, dynamic objects' states can be coupled with camera and static landmarks' states, leading to effective object tracking processes. Extensive evaluations on synthetic scenes, KITTI datasets, and real-world experiments demonstrate the performance of the method.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2021)
Proceedings Paper
Automation & Control Systems
Irene Ballester et al.
Summary: DOT is a frontend system that improves the accuracy and robustness of existing SLAM systems in highly dynamic environments by combining instance segmentation and multi-view geometry to track dynamic objects.
2021 IEEE INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION (ICRA 2021)
(2021)
Article
Automation & Control Systems
Linhui Xiao et al.
ROBOTICS AND AUTONOMOUS SYSTEMS
(2019)
Article
Robotics
Shichao Yang et al.
IEEE TRANSACTIONS ON ROBOTICS
(2019)
Article
Computer Science, Artificial Intelligence
Jakob Engel et al.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2018)
Article
Biochemical Research Methods
Balaji Veeramani et al.
BMC BIOINFORMATICS
(2018)
Proceedings Paper
Computer Science, Artificial Intelligence
Fangwei Zhong et al.
2018 IEEE WINTER CONFERENCE ON APPLICATIONS OF COMPUTER VISION (WACV 2018)
(2018)
Article
Robotics
Berta Bescos et al.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2018)
Article
Computer Science, Artificial Intelligence
Vijay Badrinarayanan et al.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2017)
Article
Robotics
Raul Mur-Artal et al.
IEEE TRANSACTIONS ON ROBOTICS
(2017)
Proceedings Paper
Computer Science, Artificial Intelligence
Keisuke Tateno et al.
30TH IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR 2017)
(2017)
Proceedings Paper
Computer Science, Artificial Intelligence
Kaiming He et al.
2017 IEEE INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV)
(2017)
Article
Robotics
Raul Mur-Artal et al.
IEEE TRANSACTIONS ON ROBOTICS
(2015)
Article
Robotics
A. Geiger et al.
INTERNATIONAL JOURNAL OF ROBOTICS RESEARCH
(2013)