Robot Cleaner Gesture Control With Event-Triggered Arm Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing robot cleaners require significant resources for real-time optical flow analysis and skin color detection, leading to unnecessary gesture detection and increased power consumption, even when no command is intended by the user.
Innovation Solution
A robot cleaner with a camera that acquires black and white images, uses an arm angle determination algorithm to identify the user's gesture by comparing image contours with stored reference information, and activates only when a gesture recognition command is detected, allowing for efficient gesture recognition and reduced power usage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If real-time optical flow analysis and skin color detection are performed continuously, then gesture recognition capability is improved, but resource consumption and power usage increase significantly
Solution Approach 1:
The robot cleaner performs gesture recognition processing periodically or event-driven rather than continuously. The processing is triggered when the robot is stationary or when specific conditions are met, rather than running optical flow analysis and skin color detection at all times, thereby reducing overall power consumption while maintaining gesture recognition capability.
Solution Approach 2:
The system pre-loads reference information for gestures into memory before actual gesture recognition is needed. This allows the robot to quickly compare captured images against pre-stored reference patterns without performing complex real-time analysis, reducing processing resources and power usage during operation.
2Measurement precision
If continuous gesture detection algorithm is executed, then gesture recognition accuracy is improved, but embedded system resource consumption increases
Solution Approach 1:
The system stores reference copies of gesture images in memory during the preliminary action phase. Instead of performing complex real-time optical flow analysis and skin color detection, the system captures an image and compares it against pre-stored reference gesture patterns, significantly reducing computational resources while maintaining recognition accuracy.
Solution Approach 2:
The patent extracts only the essential gesture recognition function from continuous video processing. By isolating the gesture detection task and using simplified image comparison against reference patterns rather than full optical flow analysis, the system reduces embedded resource consumption while preserving the core gesture recognition capability.
3Adaptability or versatility
If real-time skin color detection is performed, then gesture detection capability is improved, but false detection occurs when user does not make gestures
Solution Approach 1:
The system pre-stores reference information for valid gestures in memory before operation. This allows the robot to compare captured images against known gesture patterns, ensuring that only recognized gestures trigger detection responses and reducing false detections when no gesture is intended.
Data Source
AI summary
A robot cleaner and a method for controlling a robot cleaner are provided. The robot cleaner may include a casing, a drive disposed in the casing, a camera disposed in the casing to acquire an image of a user's gesture, and a controller that extracts an image including a user's arm image from the image acquired by the camera to determine an angle and a direction of the arm expressed by the user's gesture from the arm image and determine an intension expressed by the determined arm angle, and controls the drive based on the determined intension. The method may include receiving a voice command through a voice recognition device disposed in the robot cleaner; determining, by a controller, whether the received voice command is a gesture recognition command; acquiring an image of a user's gesture by a camera, when the received voice command is the gesture recognition command; determining, by the controller, an angle and direction of an arm expressed by the user's gesture from the image acquired by the camera; and determining, by the controller, a function intended by the determined arm angle to control a movement device so as to execute the determined function.