A SLAM system based on laser inertia real-time dynamic object removal

By combining IMU pre-integration and grid projection with a weighted optimization strategy, dynamic objects are removed in real time, solving the problem of inaccurate pose estimation in the LiDAR SLAM system and improving the accuracy and purity of localization and mapping.

CN115861481BActive Publication Date: 2026-07-03HARBIN INST OF TECH +1
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
HARBIN INST OF TECH
Filing Date
2022-12-23
Publication Date
2026-07-03

AI Technical Summary

Technical Problem

Existing LiDAR SLAM systems suffer from inaccurate pose estimation when dealing with dynamic objects, leading to decreased localization and mapping accuracy. Furthermore, existing post-processing methods cannot effectively improve the accuracy of pose output.

Method used

The algorithm employs an IMU pre-integration module, a preprocessing module, a dynamic removal module, and a back-end optimization module. It identifies dynamic feature points through raster projection and occupancy descriptors, and removes dynamic objects by combining a weighted optimization strategy, thereby improving the accuracy of pose estimation.

Benefits of technology

It effectively removes the influence of dynamic objects, improves the localization and mapping accuracy of the SLAM system, and ensures the accuracy of pose output and the purity of the map.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN115861481B_ABST
    Figure CN115861481B_ABST
Patent Text Reader

Abstract

This invention relates to a laser-based inertial real-time dynamic object removal SLAM system, specifically for robotic SLAM. Addressing the issue in existing technologies where dynamic objects are not removed before pose estimation, leading to inaccurate pose output, this application removes dynamic objects first, followed by feature matching for localization. This solves the problem of afterimages left by dynamic objects during mapping in laser SLAM systems, reducing the impact of dynamic objects on localization accuracy and improving the accuracy of localization and mapping. This application uses a vertical voxel height descriptor to describe the occupancy of dynamic objects and uses IMU pre-integration as the initial pose estimate to prioritize the removal of dynamic objects. Then, a weighted optimization strategy is employed to obtain the optimal pose estimate, mitigating the influence of dynamic objects and improving the accuracy of pose output.
Need to check novelty before this filing date? Find Prior Art