全文下载:
202504020.pdf
文章编号: 1672-6987(2025)04-0144-07DOI: 10.16351/j.1672-6987.2025.04.020
高煜昕, 柴文楠, 林凡愉, 李庆党*(青岛科技大学 自动化与电子工程学院, 山东 青岛 266061)
摘要: 为解决受IMU累积误差和状态传播的影响,导致视觉SLAM系统存在尺度漂移的问题,提出了一种基于行人航位推算(PDR)和视觉SLAM的室内定位算法。该方法将PDR计算的速度作为外部观测量集成到视觉SLAM系统中,构建尺度优化模型,减小漂移误差,修正已估计轨迹。实验结果表明:经PDR尺度优化后的视觉SLAM可以降低累计误差,并且在长期和短期的室内定位过程中具备较高精度和较好的稳定性。
关键词: 行人航位推算; 即时定位与地图构建; 尺度优化; 室内定位
中图分类号: TP 242 文献标志码: A
引用格式: 高煜昕, 柴文楠, 林凡愉, 等. 基于PDR与视觉SLAM的室内定位算法[J]. 青岛科技大学学报(自然科学版), 2025, 46(4): -.
GAO Yuxin, CHAI Wennan, LIN Fanyu, et al.An indoor positioning method based on PDR and VI-SLAM[J].Journal of Qingdao University of Science and Technology(Natural Science Edition), 2025, 46(4): -.
An Indoor Positioning Method Based on PDR and VI-SLAM
GAO Yuxin, CHAI Wennan, LIN Fanyu, LI Qingdang(College of Automation and Electronic Engineering, Qingdao University of Science and Technology, Qingdao 266061, China)
Abstract: In order to solve the problem of scale drift in visual-inertial SLAM system due to the influence of IMU cumulative error and state propagation.We propose an indoor positioning method based on pedestrian dead reckoning (PDR) and VI-SLAM. This method integrates the velocity calculated by PDR as an external observation into the VI-SLAM system, builds a scale optimization model, reduces the scale drift error of the VI-SLAM system, and corrects the estimated trajectory. The experimental results show that the VI-SLAM optimized by PDR scale can reduce the cumulative error, and have higher accuracy and better stability in the long-term and short-term indoor positioning process.
Key words: pedestrian dead reckoning; simultaneous localization and mapping; scale optimization; indoor positioning
收稿日期: 2024-08-18
基金项目: 泰山学者建设工程专项(2021-2025).
作者简介: 高煜昕(1997—), 男, 硕士研究生. * 通信联系人.