A truck cabin contour mapping and positioning method based on lidar
Through multiple scans and point cloud processing, a point cloud map of the truck compartment is constructed and local reference coordinate system matching is achieved. Combined with nonlinear optimization, the problem of precise positioning of the robot in the truck compartment is solved, achieving fast and accurate positioning effects.
Patent Information
- Application Number
- CN202211687285.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-12-27
- Publication Date
- 2025-09-12
- Estimated Expiration
- 2042-12-27
AI Technical Summary
Existing lidar positioning technology suffers from inaccurate positioning inside truck cabins due to the unstable parking position of the truck and the lack of positioning markers. This is especially true when the interior walls of the cabin are wrinkled and uneven, making it difficult to achieve accurate robot mapping and positioning.
By repeatedly scanning the truck compartment contour, collecting multi-frame point cloud data, performing point cloud filtering, region growing and contour interpolation processing, a truck compartment point cloud map is constructed, and the robot's precise positioning is achieved by constructing a local reference coordinate system and nonlinear optimization.
The consistency of the point cloud area density of the interior contour of the vehicle compartment is improved, the accuracy of the linear pose is ensured, and the robot can be quickly and accurately positioned inside the truck compartment, solving the problem of inaccurate positioning due to the unstable parking position of the truck and the lack of positioning markers.