Robot Locating Using Image-Ranging Fusion to Detect Hijacking Events
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing robot locating methods, particularly using laser distance sensors (LDS), suffer from low accuracy and fail to detect hijacking events, leading to reduced cleaning efficiency due to relocation errors.
Innovation Solution
Employ a method utilizing an image collection unit and ranging unit to match current and historical image data, determining historical pose information, and using a hill-climbing algorithm to select accurate current target pose information, enhancing detection of hijacking events and improving locating accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If laser distance sensors are used for robot locating, then the device complexity is reduced, but the measurement precision and reliability of location determination deteriorate
Solution Approach 1:
The patent combines multiple sensing modalities (laser ranging, image collection, orientation sensing) into a unified locating system. The robot integrates a ranging unit for distance measurement, an image collection unit for visual data, and an orientation unit for directional information, merging these separate functions into a coordinated locating apparatus that cross-validates data from multiple sources to improve measurement precision while maintaining manageable device complexity
Solution Approach 2:
The patent introduces image data as an intermediary element that mediates between the ranging data and the final location determination. The image collection unit captures visual information of the surrounding environment, which serves as an intermediary reference to verify and refine the location estimates derived from laser ranging, thereby improving measurement precision without requiring a complete redesign of the locating system
2Loss of information
If single pose information is used for robot locating, then the loss of information is reduced, but the reliability of location determination deteriorates due to inability to detect hijacking events
Solution Approach 1:
The patent performs preliminary matching between historical image data and current image data before finalizing the location determination. By pre-comparing image sequences and identifying consistent visual features, the system establishes a baseline of expected visual patterns that serves as a reference for detecting anomalies such as hijacking events, thereby improving reliability without significantly increasing information loss
Solution Approach 2:
The patent implements a feedback mechanism where the robot continuously compares current image data with historical image data and adjusts its location estimate accordingly. The system uses the matched historical pose information as feedback to validate current location determinations, and when discrepancies are detected (indicating potential hijacking), it can trigger re-localization or alert mechanisms, thereby enhancing reliability while maintaining efficient information usage
3Measurement precision
If image data matching is performed for robot locating, then the measurement precision and reliability of location determination improve, but the loss of time and device complexity increase
Solution Approach 1:
The patent performs partial image matching by focusing on key visual features and landmarks rather than comparing entire image datasets. The system identifies and matches salient features such as distinctive objects, wall corners, or furniture pieces in the environment, performing sufficient (but not exhaustive) comparison to achieve reliable location determination within acceptable time constraints, thereby reducing time loss while maintaining measurement precision
Solution Approach 2:
The patent segments the image matching process into distinct stages: feature extraction from current and historical images, feature matching between the two sets, and pose estimation based on matched features. This segmentation allows the system to process image data in manageable chunks rather than as a monolithic operation, reducing computational time and complexity while preserving measurement precision through systematic feature-by-feature comparison
4Loss of information
If multiple current possible pose information is generated, then the loss of information is reduced, but the device complexity and computational burden increase
Solution Approach 1:
The patent employs dynamic pose estimation by generating multiple possible pose information candidates and then dynamically selecting the most likely one based on image matching results. The system maintains a set of potential poses rather than committing to a single static estimate, allowing it to adaptively choose the best match as new image data becomes available, thereby reducing information loss while managing device complexity through dynamic rather than static processing
Data Source
AI summary
A locating method and apparatus for a robot, and a computer-readable storage medium. The locating method includes: determining current possible pose information of the robot according to current ranging data collected by a ranging unit; determining, according to first current image data collected by an image collection unit, first historical image data matching with the first current image data, the first historical image data being collected by the image collection unit at a historical moment; obtaining first historical pose information of the robot at a moment when the first historical image data is collected; and in response to quantity of the current possible pose information being at least two pieces, matching the first historical pose information with each piece of the current possible pose information, and using matched current possible pose information as current target pose information.