Robot Object Recognition via Self-Service Registration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional robots are unable to recognize non-registered objects in their activity space, leading to inconvenience and the need for complex procedures to identify such objects.
Innovation Solution
A robot equipped with an object-presence decision unit and a human-robot interface (HRI) module that determines the presence of non-registered objects through image and distance data, communicates with users to register these objects, and stores the information in a database.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the robot uses pre-stored characteristic information to recognize objects, then registered objects can be recognized accurately, but non-registered objects cannot be recognized at all
Solution Approach 1:
The robot performs self-learning by automatically detecting non-registered objects through multiple object area detectors, determining their presence, and registering them in the database without human intervention. The system uses the robot's own sensors and processing units to expand its recognition capability autonomously.
Solution Approach 2:
The system implements a feedback mechanism where the object-presence decision unit continuously monitors the photographing area, identifies non-registered objects, and triggers the object registration unit to add them to the database. This closed-loop process enables the robot to learn from its environment and improve its recognition capability over time.
2Adaptability or versatility
If the robot conducts complicated procedures to recognize non-registered objects, then it can identify unknown objects, but the process becomes complex and time-consuming
Solution Approach 1:
The object-presence decision unit is divided into multiple independent object area detectors, each responsible for detecting candidate objects using different methods (distance distribution, image characteristics, edge detection). This segmentation allows parallel processing and simplifies the overall detection procedure while maintaining comprehensive recognition capability.
Solution Approach 2:
The system performs preliminary actions by pre-defining multiple detection methods and criteria before encountering non-registered objects. The object area detectors are pre-configured with various detection algorithms, and the object registration unit is ready to immediately register newly detected objects, eliminating the need for complex real-time decision-making procedures.
3Adaptability or versatility
If the robot registers objects manually through user communication, then non-registered objects can be added to the database, but the process requires user intervention and time
Solution Approach 1:
The robot autonomously performs object registration by detecting non-registered objects through multiple detectors, determining their presence, and automatically adding them to the database without requiring user communication or intervention. This self-service capability eliminates time loss associated with manual registration processes.
Solution Approach 2:
The object-presence decision unit operates continuously to detect and register non-registered objects as they appear in the photographing area. This continuous operation ensures that the robot's recognition capability is constantly updated and expanded without interruption or delay, maintaining uninterrupted useful action.
Data Source
AI summary
A robot and a method of controlling the same are disclosed. The robot determines a presence or absence of a non-registration object if a registered object is not present in a photographing area, obtains an object name by communicating with a user if the presence of the non-registration object is decided, and additionally registers the object name in a database. Therefore, the robot recognizes the non-registration object present in the photographing area.