Robot Visual Tracking Mode Switching for Real-Time Localization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Point feature-based SLAM algorithms in visual-inertial odometry consume high computational load, reducing the real-time performance of robot localization and navigation.
Innovation Solution
A visual tracking method for robots using a window-based and projection-based matching mode alternation, where the robot switches between these modes based on tracking success or failure, and incorporates inertial data for adaptive image tracking.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If point feature-based SLAM algorithm is used, then feature tracking and image composition can be completed, but computational load increases and real-time performance decreases
Solution Approach 1:
The patent divides the tracking process into two distinct modes: projection-based matching mode for initial tracking and window-based matching mode for refined tracking. This segmentation allows the system to use the computationally lighter projection-based method first, then switch to the more accurate but heavier window-based method only when necessary, thus reducing overall computational load while maintaining tracking accuracy.
Solution Approach 2:
The patent implements dynamic switching between two tracking modes based on tracking success or failure. The system adapts its processing strategy in real-time: using projection-based matching when tracking succeeds (lower computational load) and window-based matching when tracking fails (higher accuracy). This dynamic adaptation resolves the contradiction by adjusting the balance between speed and accuracy according to actual tracking conditions.
2Measurement precision
If window-based matching mode is used continuously, then tracking accuracy is maintained, but computational load and tracking time increase
Solution Approach 1:
The patent applies partial action by using the computationally intensive window-based matching mode only partially - specifically, only when the projection-based matching mode fails. In normal successful tracking scenarios, the system uses the lighter projection-based method, thus avoiding unnecessary computational expenditure and reducing overall tracking time while maintaining accuracy when needed.
Solution Approach 2:
The projection-based matching mode serves as an intermediary that handles the majority of tracking cases with lower computational cost. When this intermediary method succeeds, the system avoids the heavier window-based matching. Only when the intermediary fails does the system resort to the full window-based matching, thus reducing tracking time while preserving accuracy through the intermediary's filtering effect.
3Productivity
If projection-based matching mode is used, then computational load is reduced, but tracking reliability decreases when tracking fails
Solution Approach 1:
The patent implements a cushioning mechanism by having the window-based matching mode ready as a backup when projection-based matching fails. This prior preparation of an alternative method ensures that tracking reliability is maintained - when the primary projection-based method encounters failure, the system can immediately switch to the more robust window-based method, cushioning against complete tracking failure.
Solution Approach 2:
The system uses feedback from the projection-based matching results to control switching to window-based matching. When projection-based matching fails (feedback signal), the system activates window-based matching to recover tracking. This feedback mechanism ensures that computational efficiency is maintained during successful tracking while reliability is preserved by activating the more robust method only when needed, based on actual tracking performance feedback.
Data Source
AI summary
Disclosed is a visual tracking method of robot, wherein an execution body of the visual tracking method of the robot is a robot fixedly equipped with a camera and an inertial sensor; the visual tracking method of the robot includes: the robot performing image tracking using a window-based matching mode, and when the robot succeeds in performing tracking using the window-based matching mode, the robot stopping performing image tracking using the window-based matching mode, and then the robot performing image tracking using a projection-based matching model; and then, when the robot fails to perform tracking using the projection-based matching mode, the robot stopping performing image tracking using the projection-based matching mode, and then the robot performing image tracking using the window-based matching mode.