Robot Positioning via Sensor Fusion of Odometry and Visual Registration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current robotic scanning and positioning systems face accuracy issues due to reliance on camera-to-robot calibration and nominal kinematics, particularly in feature-less or feature-sparse scenes, leading to distance drift and challenges in real-time synchronization between camera image acquisition and robot position reading.
Innovation Solution
A method that combines robot position measurement (robot odometry) with image registration (visual odometry) to automate camera-to-robot calibration, using sensor fusion to increase positioning accuracy by recording and synchronizing camera and robot position coordinates with timestamps, and estimating camera tool center points and scene work objects across different scan regions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If camera-to-robot calibration and nominal kinematics are used for positioning, then the system structure remains simple, but positioning accuracy deteriorates due to distance drift and error accumulation
Solution Approach 1:
The patent combines robot odometry (measured robot positions) with visual odometry (image registration methods) into a unified sensor fusion approach. This merging allows the system to leverage the advantages of both methods while compensating for their individual weaknesses, thereby improving positioning accuracy without requiring complex separate calibration systems
Solution Approach 2:
The system performs automated camera-to-robot calibration by utilizing the robot's own motion data and captured images. The calibration process is self-contained, using the robot's measured positions and visual features from the environment to automatically determine the camera tool center point, eliminating the need for external calibration equipment or procedures
2Ease of manufacture
If image registration methods are used alone, then calibration equipment is not needed, but positioning accuracy deteriorates in feature-less or feature-sparse scenes
Solution Approach 1:
The patent merges robot odometry data with visual odometry results, creating a hybrid positioning system. The robot's measured positions provide reliable distance information that compensates for the weaknesses of image registration in feature-sparse environments, while visual features help correct drift in the odometry data
Solution Approach 2:
The patent introduces an automated calibration process as an intermediary that connects the robot's motion data with the visual navigation system. This calibration mediator translates between the robot's coordinate system and the camera's coordinate system, enabling accurate positioning without requiring external calibration equipment
3Measurement precision
If separate calibration steps are performed, then camera-to-robot relation is established, but time consumption increases and real-time synchronization becomes challenging
Solution Approach 1:
The system performs self-calibration by utilizing data already collected during normal operation. The robot's measured positions and captured images are used to automatically determine the camera tool center point, eliminating the need for separate calibration steps and enabling real-time calibration without external equipment
Solution Approach 2:
The patent performs calibration continuously in the background during normal operation rather than requiring a separate preliminary calibration step. By accumulating and processing data as the robot moves, the system prepares the calibration information in advance, enabling real-time synchronization without interrupting the robot's workflow
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
Robot positioning is facilitated by obtaining, for each time of a first sampling schedule, a respective indication of a pose of a camera system of a robot relative to a reference coordinate frame, the respective indication of the pose of the camera system being based on a comparison of multiple three-dimensional images of a scene of an environment, the obtaining providing a plurality of indications of poses of the camera system; obtaining, for each time of a second sampling schedule, a respective indication of a pose of the robot, the obtaining providing a plurality of indications of poses of the robot; and determining, using the plurality of indications of poses of the camera system and the plurality of indications of poses of the robot, an indication of the reference coordinate frame and an indication of a reference point of the camera system relative to pose of the robot.