针对图书馆室内导航机器人惯性导航定位不准、导航有误差的问题,文章基于机器人操作系统2(Robot Operating System 2,ROS2),采用2D激光雷达(Light Detection and Ranging,LiDAR)与惯性测量单元(Inertial Measurement Unit,IMU)相结合的...针对图书馆室内导航机器人惯性导航定位不准、导航有误差的问题,文章基于机器人操作系统2(Robot Operating System 2,ROS2),采用2D激光雷达(Light Detection and Ranging,LiDAR)与惯性测量单元(Inertial Measurement Unit,IMU)相结合的方案,对图书馆室内移动机器人进行建模,并在此基础上对比Cartographer、Gmapping两种同步定位与地图构建(Simultaneous Localization and Mapping,SLAM)技术的建图效果,通过自适应蒙特卡洛定位(Adaptive Monte Carlo Localization,AMCL)算法定位,完成A*和Dijkstra导航算法的比较,最终实现了机器人SLAM与路径规划仿真,验证了基于ROS2实现图书馆室内移动机器人导航仿真的可行性。展开更多
The current inertial measurement unit(IMU)and odometry fusion navigation algorithms often incorporate non-holonomic constraints(NHC)to obtain three-dimensional velocity in the navigation frame.However,due to the integ...The current inertial measurement unit(IMU)and odometry fusion navigation algorithms often incorporate non-holonomic constraints(NHC)to obtain three-dimensional velocity in the navigation frame.However,due to the integral nature of the dead reckoning algorithm,the attitude errors of the IMU accumulate over time,causing the velocity transformation results to fail to accurately reflect the threedimensional velocity in the navigation frame.Based on the fact that during a vehicle's horizontal and uniform motion,the vertical acceleration is consistent with gravitational acceleration,this paper proposes an IMU/odometry fusion navigation algorithm based on horizontal attitude constraints(HAC).Building on non-holonomic constraints,this algorithm determines the motion state of the vehicle through accelerometer output and zeroes out the pitch and roll angles during horizontal and uniform motion.Verified through two sets of real-world vehicle test data,this algorithm improves horizontal positioning accuracy by approximately 63%and 70%,and vertical positioning accuracy by 98%and 97%,compared with the traditional NHC IMU/odometer fusion algorithm.展开更多
Human pose estimation is crucial across diverse applications,from healthcare to human-computer interaction.Integrating inertial measurement units(IMUs)with monocular vision methods holds great potential for leveraging...Human pose estimation is crucial across diverse applications,from healthcare to human-computer interaction.Integrating inertial measurement units(IMUs)with monocular vision methods holds great potential for leveraging complementary modalities;however,existing approaches are often limited by IMU drift,noise,and underutilization of visual information.To address these limitations,we propose a novel dual-stream feature extraction framework that effectively combines temporal IMU data and single-view image features for improved pose estimation.Short-term dependencies in IMU sequences are captured with convolutional layers,while a Transformerbased architecture models long-range temporal dynamics.To mitigate IMU drift and inter-sensor inconsistencies,a complementary filtering module is introduced alongside a cross-channel interaction mechanism.Features from the IMU and image streams are then fused via a dedicated fusion module and further refined utilizing a high-precision regression head for accurate pose prediction.Experimental results on benchmark datasets demonstrate that our method significantly outperforms existing techniques in terms of estimation,accuracy,and robustness,validating the effectiveness of our dual-stream architecture.展开更多
文摘针对图书馆室内导航机器人惯性导航定位不准、导航有误差的问题,文章基于机器人操作系统2(Robot Operating System 2,ROS2),采用2D激光雷达(Light Detection and Ranging,LiDAR)与惯性测量单元(Inertial Measurement Unit,IMU)相结合的方案,对图书馆室内移动机器人进行建模,并在此基础上对比Cartographer、Gmapping两种同步定位与地图构建(Simultaneous Localization and Mapping,SLAM)技术的建图效果,通过自适应蒙特卡洛定位(Adaptive Monte Carlo Localization,AMCL)算法定位,完成A*和Dijkstra导航算法的比较,最终实现了机器人SLAM与路径规划仿真,验证了基于ROS2实现图书馆室内移动机器人导航仿真的可行性。
基金from the National Key Research and Development Program project"Adaptive Navigation Software and Hardware Technology(2018YFB0505200)."。
文摘The current inertial measurement unit(IMU)and odometry fusion navigation algorithms often incorporate non-holonomic constraints(NHC)to obtain three-dimensional velocity in the navigation frame.However,due to the integral nature of the dead reckoning algorithm,the attitude errors of the IMU accumulate over time,causing the velocity transformation results to fail to accurately reflect the threedimensional velocity in the navigation frame.Based on the fact that during a vehicle's horizontal and uniform motion,the vertical acceleration is consistent with gravitational acceleration,this paper proposes an IMU/odometry fusion navigation algorithm based on horizontal attitude constraints(HAC).Building on non-holonomic constraints,this algorithm determines the motion state of the vehicle through accelerometer output and zeroes out the pitch and roll angles during horizontal and uniform motion.Verified through two sets of real-world vehicle test data,this algorithm improves horizontal positioning accuracy by approximately 63%and 70%,and vertical positioning accuracy by 98%and 97%,compared with the traditional NHC IMU/odometer fusion algorithm.
基金support provided by the European University of Atlantic.
文摘Human pose estimation is crucial across diverse applications,from healthcare to human-computer interaction.Integrating inertial measurement units(IMUs)with monocular vision methods holds great potential for leveraging complementary modalities;however,existing approaches are often limited by IMU drift,noise,and underutilization of visual information.To address these limitations,we propose a novel dual-stream feature extraction framework that effectively combines temporal IMU data and single-view image features for improved pose estimation.Short-term dependencies in IMU sequences are captured with convolutional layers,while a Transformerbased architecture models long-range temporal dynamics.To mitigate IMU drift and inter-sensor inconsistencies,a complementary filtering module is introduced alongside a cross-channel interaction mechanism.Features from the IMU and image streams are then fused via a dedicated fusion module and further refined utilizing a high-precision regression head for accurate pose prediction.Experimental results on benchmark datasets demonstrate that our method significantly outperforms existing techniques in terms of estimation,accuracy,and robustness,validating the effectiveness of our dual-stream architecture.
