Mobile Robot Localization Using Direction Descriptors and Mode Switching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current mobile robot navigation and localization methods face challenges with significant illumination changes and image variations, leading to errors in feature point matching and adaptive driving control, especially when inertial sensor data is unreliable.
Innovation Solution
A control method for mobile robots that employs local direction descriptors, using image processing units to extract feature points, generate descriptors, and switch between driving modes based on image quality and matching degrees, allowing for adaptive control through image and sensing information integration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional feature point descriptors are used for mobile robot navigation, then the system can process images, but the performance deteriorates when illumination changes significantly or image variations occur
Solution Approach 1:
The patent transforms image patches into frequency domain representations using Fourier transforms, changing the parameter space from spatial domain to frequency domain. This transformation makes the descriptors invariant to illumination changes and rotational variations, as frequency domain representations capture essential structural information while filtering out intensity variations.
Solution Approach 2:
The patent replaces traditional mechanical/optical feature detection methods with frequency domain analysis. Instead of relying on intensity-based feature detection that is sensitive to illumination, the system uses spectral analysis to extract rotationally invariant and illumination-resistant descriptors.
2Measurement precision
If inertial sensor data is used for driving control, then the robot can navigate using sensor fusion, but wrong recognition occurs when sensor errors accumulate
Solution Approach 1:
The patent implements a feedback mechanism where image-based feature point matching results are used to correct and verify inertial sensor-derived position estimates. When feature point matching succeeds, it provides feedback to reset accumulated drift errors from inertial sensors, creating a hybrid navigation system that leverages the strengths of both sensing modalities.
Solution Approach 2:
The system prepares multiple candidate feature point matches and pre-validates them using geometric constraints and consistency checks before accepting position corrections. This beforehand validation cushions against wrong recognition errors by filtering out incorrect matches before they can corrupt the navigation state.
3Adaptability or versatility
If adaptive driving control is implemented, then the robot can respond to different conditions, but the system complexity increases when multiple driving modes are required
Solution Approach 1:
The patent implements dynamic switching between different driving modes based on real-time assessment of image quality, feature point matchability, and sensor reliability. The system transitions from inertial-sensor-dependent navigation to image-based navigation when illumination changes or sensor drift is detected, providing adaptability without requiring permanently complex control structures.
Data Source
Figure 1
Figure 2(a)~2(b)
Figure 3
AI summary
The present invention relates to a control method for the localization and navigation of a mobile robot and a mobile robot using the same. More specifically, the localization and navigation of a mobile robot are controlled using inertial sensors and images, wherein local direction descriptors are employed, the mobile robot is changed in the driving mode thereof according to the conditions of the mobile robot, and errors in localization may be minimized.