Moving Robot Ceiling-Based Map Correction for Area Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing moving robots generate maps that do not accurately reflect the shape of indoor spaces due to obstacles, leading to difficulties in user recognition and correct command input, especially in complex environments with multiple connected areas.
Innovation Solution
A moving robot system that captures images of ceilings to extract features, filters unnecessary information, and combines feature correlations to generate maps that reflect the actual shape of areas, including obstacles, allowing for accurate representation of travelable spaces.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the moving robot generates a map based on obstacle detection during traveling, then the map reflects the travelable area, but the map shape does not match the user's recognition of the indoor space
Solution Approach 1:
The patent combines two different mapping approaches: obstacle-based mapping (which accurately reflects travelable areas) and ceiling image-based mapping (which reflects the user's perception of indoor space boundaries). By merging these two maps, the system produces a final map that satisfies both accuracy requirements and user recognition needs.
Solution Approach 2:
The patent introduces ceiling images as an intermediary element to bridge the gap between obstacle detection data and user perception. The ceiling images serve as a mediator that captures the user's view of the indoor space boundaries, which then guides the correction of the obstacle-based map to match user expectations.
2Reliability
If the moving robot uses obstacle detection to generate the map, then the map shows travelable spaces, but the complexity of the map increases when multiple indoor spaces are connected
Solution Approach 1:
The patent extracts the boundary information of indoor spaces from ceiling images, separating the structural boundary definition from the obstacle detection process. This extraction allows the system to simplify complex multi-space maps by using ceiling-derived boundaries to organize and structure the overall layout, reducing visual and conceptual complexity.
Solution Approach 2:
The patent segments the mapping process into two independent components: obstacle-based travelable area detection and ceiling-based boundary detection. This segmentation allows each component to focus on its specific function, making the overall system more manageable and the resulting map easier to interpret despite complex environments.
3Productivity
If the moving robot generates a map without ceiling shape correction, then the generation process is simpler, but the user cannot easily divide areas or input correct cleaning commands
Solution Approach 1:
The patent performs preliminary action by capturing ceiling images and extracting boundary information during the robot's initial traversal of the space. This preliminary extraction of spatial structure information is done before final map generation, allowing the subsequent map creation and command input processes to proceed more efficiently with pre-organized boundary data.
Data Source
Figure 1~2
Figure 3(a)~3(b)
Figure 4~5(e)
AI summary
According to a moving robot and a method of controlling the same of the present disclosure, a shape of a ceiling is extracted from an image captured during traveling on the basis of a map generated by detecting an obstacle, a shape of an area corresponding to the shape of the ceiling is extracted, an actual shape of the area and a map to which a detection result of the obstacle is reflected are generated, and a map having a shape close to an actual shape of an indoor shape is provided to a user. Accordingly, the user can easily divide the plurality of areas, can input a correct cleaning command based on the map, and can easily check a current state or a cleaning state of the moving robot for each area.