Feedback-Guided Object Rotation for Robot Vision
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques for object recognition by robots are limited by the restricted movable range of the robot, which can restrict the acquisition of multi-viewpoint images and accurate feature determination of unknown objects, necessitating user intervention to change the object's attitude.
Innovation Solution
An information processing device that detects a hand region in an input image, identifies and specifies an unknown object candidate region, determines if the object is unknown, and generates feedback information to encourage the user to change the object's attitude, allowing for accurate feature extraction and recognition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the robot rotates the object to acquire multi-viewpoint images, then the robot can obtain images from different angles, but the contents of the acquired images are restricted by the robot's movable range
Solution Approach 1:
Instead of the robot rotating the object, the system inverts the approach by having the user rotate the object based on feedback from the information processing device. This allows viewpoints beyond the robot's movable range to be captured while maintaining effective object recognition.
Solution Approach 2:
The system introduces feedback information as an intermediary between the object recognition system and the user. This feedback mechanism guides the user to rotate the object to optimal viewpoints, effectively bridging the gap between limited robot mobility and comprehensive viewpoint acquisition.
2Extent of automation
If the robot autonomously rotates the object, then the process is automated, but the robot's restricted movable range limits the quality and variety of acquired images
Solution Approach 1:
The system implements a feedback loop where the information processing device analyzes the object's current姿态 and provides guidance feedback to the user. This feedback mechanism enables semi-automation, combining automated analysis with human capability to achieve both high automation level and comprehensive viewpoint acquisition.
Solution Approach 2:
The system enables the user to perform the rotation task themselves based on system feedback, making the user serve their own purpose in the recognition process. This approach leverages human flexibility to overcome robotic mobility limitations while maintaining system automation.
3Measurement precision
If the system requires comprehensive multi-viewpoint images for unknown object recognition, then recognition accuracy improves, but the process becomes more complex and time-consuming
Solution Approach 1:
The information processing device provides real-time feedback to the user about which viewpoints have been captured and which are still needed. This feedback-driven approach simplifies the complex process of acquiring comprehensive multi-viewpoint images by guiding users through a structured sequence of actions rather than requiring them to understand the full complexity of viewpoint requirements.
Data Source
AI summary
An information processing device includes an identification unit that detects a hand region of a person who is in contact with an object included in an input image and identifies an object region that is a region of the object based on a result thereof, a specification unit that specifies an unknown object candidate region that is a candidate for a region of an unknown object from among the object regions identified by the identification unit, a determination unit that determines whether the object is an unknown object based on the unknown object candidate region, and a generation unit that generates, when the object is an unknown object, feedback information for changing an attitude of the unknown object with respect to the person.