Robot Localization Using Omni-Directional Image Correlation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional localization methods for moving robots using omni-directional cameras are inaccurate due to their sensitivity to errors and inability to correctly recognize the robot's location, as they estimate only approximate positions and rely on dynamic programming algorithms.
Innovation Solution
A localization method that captures omni-directional images, calculates correlation coefficients between current and stored images using Fast Fourier Transform, and determines the robot's location by identifying nodes with high correlation coefficients, adjusting the heading angle based on the number of effective nodes to accurately navigate to the correct location.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If dynamic programming algorithm is used for image matching, then the localization process can be implemented, but the localization accuracy is greatly deteriorated due to sensitivity to errors
Solution Approach 1:
The patent changes the matching algorithm from dynamic programming to a correlation coefficient-based method. This parameter change in the algorithmic approach eliminates the sensitivity to errors that plagues dynamic programming, thereby improving both reliability and measurement precision in localization.
Solution Approach 2:
The patent replaces the dynamic programming algorithm (a computational mechanical system) with a correlation coefficient calculation method. This substitution uses a different computational mechanism that is inherently more robust to errors in image matching, thus resolving the contradiction between reliability and precision.
2Ease of operation
If omni-directional image comparison is used for localization, then the robot can navigate, but only approximate location can be estimated instead of correct location
Solution Approach 1:
The patent applies preliminary action by pre-storing omni-directional images at multiple known nodes before the actual localization process. This allows the robot to compare current images against a comprehensive pre-established database, enabling precise location determination rather than just approximate estimation.
Solution Approach 2:
The patent implements feedback by continuously comparing the currently captured omni-directional image with stored images from multiple nodes and using correlation coefficients to determine the best match. This feedback mechanism refines the location estimation from approximate to precise by iteratively identifying the node with the highest correlation.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This method enables precise localization of moving robots by accurately determining the highest correlation node and adjusting the heading angle, ensuring correct recognition of the robot's location and addressing the limitations of previous methods.
Implementation Method 1
an omni-directional camera adapted to acquire an omni-directional image (i.e., an image of 360° in the vicinity of the camera)
Implementation Method 2
calculates correlation coefficients between current and stored images using Fast Fourier Transform
Data Source
AI summary
A localization method of a moving robot is disclosed in which the moving robot includes: capturing a first omni-directional image by the moving robot; confirming at least one node at which a second omni-directional image having a high correlation with the first omni-directional image is captured; and determining that the moving robot is located at the first node when the moving robot reaches a first node, at which a second omni-directional image having a highest correlation with the first omni-directional image is captured, from among the at least one node.