Virtual Robot Teaching Using Base and Finger Image Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing robot teaching systems require actual operation of the robot arm to perform teaching, even when an end effector or peripheral device is absent, which is inefficient and limiting.
Innovation Solution
A method and system that utilize a base and finger section for teaching, photographed and processed to recognize positions and postures, calculate joint angles, and display a three-dimensional robot image in a virtual space without physical robot operation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If actual robot operation is required for teaching, then teaching accuracy can be ensured, but teaching efficiency and flexibility deteriorate
Solution Approach 1:
The patent uses teaching sections (base section and finger section) as physical copies or proxies for the actual robot components. These teaching sections are photographed and processed to create virtual models, allowing teaching operations to be performed on the copies rather than the actual robot, thereby improving efficiency while maintaining accuracy through precise image recognition and coordinate transformation
Solution Approach 2:
The patent replaces the mechanical robot operation system with an image processing and virtual space system. Instead of physically moving the robot arm to teach positions, the system photographs teaching sections, recognizes their coordinates through image processing, calculates joint angles mathematically, and displays virtual robot images, substituting mechanical manipulation with optical and computational methods
2Reliability
If physical robot manipulation is required, then real robot state verification is possible, but teaching flexibility and ease of operation worsen
Solution Approach 1:
The system creates a virtual copy of the robot in virtual space that mirrors the actual robot's structure and joint configurations. Operators can verify and adjust robot states by manipulating this virtual model, which maintains reliability through accurate coordinate transformation while dramatically improving flexibility by allowing operations without physical robot presence
Solution Approach 2:
The patent introduces a virtual space and image processing system as an intermediary between the operator and the actual robot. This intermediary layer allows operators to work with virtual images and calculated joint angles rather than directly manipulating physical components, enhancing ease of operation while maintaining reliability through mathematical transformations that preserve the actual robot's kinematic relationships
Data Source
AI summary
A robot image display method includes (a) a step of recognizing the position and the posture of a base of a robot from a base section image of a base section for teaching, (b) a step of recognizing the position and the posture of a finger section of the robot from a finger section image of a finger section for teaching, (c) a step of calculating angles of one or more joints of the robot from the position and the posture of the base and the position and the posture of the finger section, and (d) a step of displaying, in a virtual space, a three-dimensional image of the robot in a state in which the joints are at the angles calculated in the step (c).