Method and device for recognizing closely placed tray poses, computer device, medium
By acquiring and processing 3D point cloud data and using image recognition technology, the problem of inaccurate pallet pose positioning in existing technologies has been solved, enabling accurate positioning of closely placed pallets and improving the accuracy of forklift operation.
CN115546300BActive Publication Date: 2026-05-29VISIONNAV ROBOTICS SHENZHEN LTD
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- VISIONNAV ROBOTICS SHENZHEN LTD
- Filing Date
- 2022-10-11
- Publication Date
- 2026-05-29
AI Technical Summary
Technical Problem
In existing technologies, 2D laser rangefinders can only obtain the length of the holes on a plane, and cannot accurately identify and locate the pose of closely placed pallets, resulting in inaccurate pallet pose positioning by forklifts.
Method used
By acquiring the original 3D point cloud data of the target pickup area, projecting multiple target points, generating the original pallet image, extracting the target point cloud data of the pallet to be picked up, calculating its pose information, and combining 3D laser scanning and image recognition technologies, the positioning accuracy is improved.
Benefits of technology
It achieves accurate positioning of the pallet to be picked up, improving the pallet recognition accuracy of forklifts in closely packed environments.
✦ Generated by Eureka AI based on patent content.
Smart Images
Figure CN115546300B_ABST
Abstract
The application relates to a closely arranged tray pose recognition method and device, computer equipment and a medium. The method comprises the following steps: acquiring original 3D point cloud data of a target picking area; wherein the target picking area is arranged with to-be-picked trays, a plurality of the to-be-picked trays are closely arranged, and the to-be-picked trays comprise a plurality of tray foot piers; a plurality of target points are selected from the original 3D point cloud data, and the plurality of target points are projected to obtain an original tray image; target point cloud data corresponding to the to-be-picked trays is extracted from the original 3D point cloud data according to the original tray image; and pose information of the to-be-picked trays is calculated according to the target point cloud data. The method can realize accurate positioning and identification of the tray pose.
Need to check novelty before this filing date? Find Prior Art