基于视觉、激光等单一传感器的定位方法难以适应多样化的环境,围绕激光雷达、惯性测量单元和相机3 种模态的传感器信息源,针对激光雷达(Light Detection and Ranging,LiDAR)与视觉测量没有充分关联的问题,提出了一种基于多状态颜色一致性约束的激光雷达-惯性-视觉里程计方法,以提高系统的鲁棒性和定位精度.该方法紧耦合了激光雷达-惯性里程计(LiDAR-Inertial Odometry,LIO)和视觉-惯性里程计(Visual-Inertial Odometry,VIO)两个子系统,并定义了带有颜色信息的全局地图表示形式.LIO子系统中点云经过运动补偿后,直接用于构建点到面的残差.VIO子系统利用全局地图中点的深度信息,根据滑动窗口中多个相机状态观测到同一地图点颜色的一致性,构建光度误差约束,并通过不变扩展卡尔曼滤波(Extended Kalman Filter,EKF)状态估计器进行系统状态更新.在南洋理工大学发布的公共数据集上进行了实验,所提方法在该数据集不同序列上的绝对轨迹误差平均值为0.402 m.
A LiDAR-Inertial-Visual Odometry Based on Multi-state Color Consistency Constraints
The localization method based on a single sensor is difficult to adapt to diverse environments.Three sensor information sources of light detection and ranging(LiDAR),inertial measurement unit and camera are focused on.For the problem that LiDAR and visual measurement can not be fully correlated,a LiDAR-inertial-visual odometry method based on multi-state color consistency constraint is proposed to improve the robustness and localization accuracy of the system.This method tightly couples the LiDAR-inertial odometry(LIO)and visual-inertial odometry(VIO)subsystems and defines a global map representation form that incorporates color information.In the LIO subsystem,the motion-compensated point cloud is directly employed to construct point-to-plane residuals for optimizing the system state.The VIO subsystem directly utilizes the depth information of points in the global map and constructs photometric error constraints based on the invariance of the color observed by multiple camera states for the same map point within the sliding window.The system state is then updated through an invariant extended Kalman filter(EKF)state estimator.Experiments are conducted on a public dataset released by Nanyang Technological University.The average absolute trajectory error of the proposed method on different sequences of the dataset is 0.402 meters.
万方数据
维普
