A semantic map-based underground and overground mapping and positioning method for a robot dog
By employing layered and progressive sensor fusion and semantic map optimization, the problems of positioning drift and insufficient information fusion in complex environments of robot dogs have been solved, achieving high-precision and robust positioning and scene understanding, and adapting to the switching between ground and underground spaces.
CN122149454APending Publication Date: 2026-06-05ZHIHAN XINGTU (SUZHOU) TECH CO LTD +1
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- ZHIHAN XINGTU (SUZHOU) TECH CO LTD
- Filing Date
- 2026-02-03
- Publication Date
- 2026-06-05
Smart Images
Figure CN122149454A_ABST
Abstract
The application provides a semantic map-based underground and overground mapping and positioning method for a robot dog, and belongs to the technical field of autonomous navigation of mobile robots; the method comprises the following steps: acquiring original observation data through sensors carried by the robot dog; unifying the original observation data to a consistent space-time reference to obtain preprocessed observation data; performing sensor data compound filtering; adopting a multi-stage series filtering strategy to smooth the preprocessed observation data after time synchronization to obtain smoothed sensor observation values; setting a repositioning condition and taking the repositioning condition as a loop constraint; adopting an error state Kalman filter to obtain a predicted position of the robot dog through calculation; and adaptively switching different sources of observation data of the surrounding environment according to the surrounding environment and the actual elevation of the predicted position of the robot dog to construct an environment map.
Need to check novelty before this filing date? Find Prior Art