Robot Calibration Using Marker Inclination to Reduce Quantization Error
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current robot system calibration methods face challenges in achieving high accuracy due to quantization errors in the optical axis direction and the need for special jigs with accurate design data, which increases costs and man-hours.
Innovation Solution
A robot system and calibration method that utilize at least three teaching points, where two points are placed on the same plane normal to the optical axis with the same inclination, and the third point has a different inclination, allowing for reduced quantization error and eliminating the need for special jigs with accurate design data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual trial and error is used to search for teaching points in the actual operation site, then the calibration process can be completed, but the calibration accuracy is insufficient due to selection of inappropriate teaching points
Solution Approach 1:
The patent changes the parameters of teaching points from arbitrary manual selection to systematically designed positions and orientations. By specifying that two teaching points should be on the same plane normal to the optical axis with the same inclination, and the third point with a different inclination, the calibration accuracy is improved while eliminating the difficulty of manual trial and error.
2Measurement precision
If a special jig with accurate design data is used for calibration, then the calibration accuracy can be improved, but the cost and man-hours increase
Solution Approach 1:
Instead of using a special jig with accurate design data, the patent uses a standard jig with markers whose positions and orientations are copied from actual operation sites. This approach achieves high calibration accuracy by reflecting real-world conditions while avoiding the high cost and long lead time of manufacturing special jigs with precise design data.
Solution Approach 2:
The patent makes a standard jig serve multiple functions: it provides both the physical structure for positioning and the reference markers for calibration. By attaching markers to a standard jig and defining their positions and orientations to match actual operation sites, the system eliminates the need for dedicated special jigs while maintaining calibration accuracy.
3Measurement precision
If teaching points are selected without considering marker inclination and position relationships, then the calibration process is simpler, but quantization errors in the optical axis direction significantly reduce measurement accuracy
Solution Approach 1:
The patent introduces specific parameter constraints for teaching point configuration: two points on the same plane normal to the optical axis with the same inclination, and one point with a different inclination. These parameter changes eliminate quantization errors in the optical axis direction by ensuring that marker positions can be accurately determined in three-dimensional space, thereby improving measurement accuracy without excessive complexity.
Data Source
AI summary
A control apparatus calculates a calibration value based on a position in the robot coordinate system 41 and a position in the vision coordinate system 42, for at least three teaching points set within a calibration area. Markers 21 of two of the three teaching points have the same inclination in relation to an optical axis of a camera 3 as a visual sensor, and the two points are placed on different positions of the same plane normal to the optical axis. The remaining one of the three teaching points other than the two points is set such that the inclination of the marker 21 in relation to the optical axis is different from that of the two points. As a result, influence of a large quantization error in the optical axis direction as a measurement error of the camera 3 can be reduced.