Robot Arm Calibration Using Stereo Visual Sensors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current robot calibration methods, such as coordinate positioning through inching and the use of calibration jigs, are time-consuming, skill-dependent, and costly, with inaccuracies arising from operator error and potential marker obstruction, especially when the detection and approach directions of the visual sensor and robot arm are similar.
Innovation Solution
A robot apparatus equipped with a fixed visual sensor and a movable visual sensor, along with a control device that calculates a calibration function to relate measurement data from both sensors, allowing for accurate calibration without fine operator adjustments and marker obstruction issues, by using a stereo camera system that can detect features within a wide range of the robot arm's movement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If coordinate positioning through inching is used for calibration, then the robot arm can be positioned to match features, but the process becomes time-consuming and operator skill-dependent
Solution Approach 1:
The patent replaces the manual mechanical inching process with an automated visual measurement system. The visual sensor automatically detects feature coordinates, and the control device calculates calibration values through coordinate transformation, eliminating the need for operator manual positioning while maintaining high precision.
Solution Approach 2:
The system performs self-calibration by automatically detecting features and computing calibration parameters without requiring operator intervention. The robot arm moves to predetermined positions, the visual sensor captures images, and the control device automatically processes the data to generate calibration values.
2Area of stationary object
If a fixed visual sensor is used for detection, then the detection range can cover the robot arm's movement area, but markers may be obstructed when detection and approach directions are similar
Solution Approach 1:
The patent introduces a second visual sensor mounted on the robot arm that provides a different spatial perspective. This additional dimension in sensor placement allows detection of markers from multiple angles, preventing obstruction issues that occur with a single fixed sensor when detection and approach directions are similar.
Solution Approach 2:
The second visual sensor acts as an intermediary detection device that captures marker positions from the robot arm's perspective. This intermediary sensor provides complementary measurement data that resolves the obstruction problem by detecting markers that may be hidden from the fixed sensor's view.
3Measurement precision
If calibration jigs are used for positioning, then the robot arm can be calibrated to feature coordinates, but accurate jigs are costly and require careful manufacturing
Solution Approach 1:
The patent replaces the physical calibration jig with a virtual coordinate system established through visual detection. The control device calculates calibration values by transforming coordinates between the visual sensor system and robot arm system, eliminating the need for expensive, precision-manufactured physical jigs while maintaining calibration accuracy.
Solution Approach 2:
Instead of using physical calibration jigs, the system creates a virtual model of the feature coordinates through visual detection. The control device copies the spatial relationships and uses them for calibration calculations, replacing the need for physical reference objects with digital coordinate representations.
Data Source
AI summary
A robot system includes a fixed camera that obtains first measurement data by detecting a plurality of features positioned within a detection range, the detection range including at least part of a range in which a robot arm is movable, a hand camera movable with the robot arm, and a control apparatus that controls the robot arm. A calibration function that relates a value obtained as part of the first measurement data to a command value provided to the robot arm at each of a plurality of positions and orientations at which the hand camera obtains second measurement data by detecting each mark.