Artificial intelligence vacuum cleaner and control method therefor
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing robot cleaners with single-layer obstacle recognition systems struggle to accurately identify various types of obstacles in a cleaning area, leading to inferior recognition accuracy.
Innovation Solution
A robot cleaner equipped with a control unit that performs a primary recognition process using a first recognition part to determine if an image corresponds to one of multiple obstacle types, followed by a secondary recognition process using a second recognition part to verify the result, thereby improving obstacle recognition accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single-layer obstacle recognition system is used, then the device complexity is low, but the obstacle recognition accuracy is insufficient
Solution Approach 1:
The obstacle recognition system is segmented into multiple layers: a first recognition part that performs initial obstacle type determination, and a second recognition part that performs verification. This segmentation allows each layer to specialize in specific recognition tasks, improving overall accuracy while managing complexity through modular design
Solution Approach 2:
The system transitions from a single-layer (one-dimensional) recognition approach to a multi-layer (multi-dimensional) recognition architecture. By adding the verification layer, the system creates a deeper processing dimension that enhances recognition accuracy without linearly increasing complexity
2Productivity
If a single-layer obstacle recognition system is used, then the device complexity is low, but the productivity of cleaning operations is reduced due to inaccurate obstacle identification
Solution Approach 1:
The recognition system is divided into specialized segments: the first recognition part handles initial classification of obstacle types, while the second recognition part focuses on verification. This segmentation enables more accurate obstacle identification, leading to better cleaning decisions and improved productivity
Solution Approach 2:
The second recognition part provides feedback verification of the first recognition part's results. This feedback mechanism ensures accurate obstacle identification before cleaning operations proceed, improving productivity by preventing errors while maintaining manageable system complexity
3Measurement precision
If a multi-layer obstacle recognition system is implemented, then the obstacle recognition accuracy is improved, but the loss of time for processing increases
Solution Approach 1:
The first recognition part performs preliminary obstacle type determination before the second recognition part conducts verification. This preliminary action allows the system to quickly classify most obstacles and only apply the time-consuming verification process when necessary, improving accuracy while minimizing time loss
Solution Approach 2:
The system applies partial verification: the second recognition part verifies results selectively rather than processing every single case through both layers. This approach achieves high accuracy for critical decisions while reducing overall processing time
Data Source
Figure 1
Figure 2~3
Figure 4~5
AI summary
In order to solve the problem of the present invention, an artificial intelligence vacuum cleaner for performing autonomous traveling, according to one embodiment of the present invention, comprises: a main body; a driving unit for moving the main body within a cleaning area; a camera for photographing an area around the main body; and a control unit for controlling, on the basis of an image captured by means of the camera, the driving unit such that a predetermined traveling mode is performed, wherein the control unit performs a first recognition process for determining whether the image corresponds to any one of multiple types of obstacles, performs a second recognition process for re-determining whether the image corresponds to any one obstacle type in order to verify the result of the first recognition process, and controls the driving unit on the basis of the obstacle type determined through the first and second recognition processes such that the main body travels in a preset pattern.