Mobile Robot Doorway Detection Using Occupancy Map Change Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing door detection methods for mobile robots are ineffective in complex environments with non-standard or partially obscured doors, and computer vision-based approaches are computationally demanding and require large data sets, limiting their applicability in real-world settings.
Innovation Solution
A method using a set of processors and sensors to generate and analyze current and prior occupancy maps, employing raycasting and machine learning to identify potential doorways by analyzing distances and angles, and using computer vision techniques to denoise and crop maps for accurate door detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If computer vision-based approaches are used to detect different types of doors, then adaptability is improved, but computational demand increases
Solution Approach 1:
The patent segments the door detection process into distinct phases: initial occupancy map generation, change detection between time points, and doorway identification. This segmentation allows the system to process only relevant changes rather than analyzing complete scenes, reducing computational demand while maintaining adaptability across different door types
Solution Approach 2:
The system performs preliminary actions by generating occupancy maps at earlier time points and storing them for later comparison. This preliminary mapping allows the robot to efficiently detect doorways by comparing changes over time rather than performing full scene analysis each moment, reducing real-time computational requirements
2Adaptability or versatility
If computer vision-based approaches are used to detect different types of doors, then adaptability is improved, but device complexity increases
Solution Approach 1:
The patent implements a universal occupancy map representation that can detect various door types (sliding doors, hinged doors, automatic doors) using the same fundamental approach. The system uses multi-functional sensors that can detect both static occupancy and temporal changes, eliminating the need for separate detection systems for different door types while maintaining high adaptability
3Device complexity
If standard geometric shape recognition is used for door detection, then device complexity is reduced, but measurement precision deteriorates
Solution Approach 1:
The patent transitions from static geometric shape recognition to a dynamic change detection approach. Instead of relying on fixed geometric features that may not match non-standard doors, the system detects doorways by identifying temporal changes in occupancy patterns, allowing accurate detection of doors with varying shapes, sizes, and configurations while maintaining relatively simple system architecture
Data Source
AI summary
Methods are disclosed that allow a robot, equipped with sensors and processors, to develop maps of its surroundings as it moves through a human environment, and to use these maps to detect and identify open doors, which it can then move through. Robots are disclosed to develop and use such maps to detect and identify open doors and move therethrough.