Robot Arm Pose Error Detection Using Visual Marker Association
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing robot systems for remote operations face challenges in accurately detecting pose errors of the operating arm in real-time, which is crucial for maintaining operation accuracy and human-computer interaction experience.
Innovation Solution
The proposed solution involves an error detection method that includes obtaining a target pose of the operating arm's end, acquiring a positioning image, recognizing pose and angle identifications on the arm, determining the actual pose based on these identifications, and generating a control signal if an error detection condition is met.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional motion conversion control is used between master manipulator and operating arm, then the robot system can achieve basic remote operation functionality, but the pose error detection accuracy is insufficient and real-time error correction capability is lacking
Solution Approach 1:
The patent introduces pose identifications (marker balls with coded patterns) as intermediary objects attached to the operating arm. These markers serve as mediators between the operating arm and the detection system, enabling precise pose measurement through image recognition without requiring complex direct sensing on the arm itself. The markers provide easily detectable visual features that bridge the gap between physical position and digital measurement.
Solution Approach 2:
The patent creates a visual copy of the operating arm's pose through 2D images captured by the camera. Instead of directly measuring 3D spatial coordinates, the system captures 2D images containing pose identifications, recognizes the marker positions and orientations in the image plane, and infers the actual 3D pose from this visual copy. This copying approach simplifies the measurement process while maintaining accuracy.
2Reliability
If real-time pose error detection is implemented, then operation accuracy and safety are improved, but the system response time and computational load increase
Solution Approach 1:
The patent pre-attaches pose identifications (marker balls with known patterns) to the operating arm before operation begins. These markers are in fixed positions with known geometric relationships, so when the camera captures an image, the system can immediately recognize them and calculate pose errors without needing to search for or identify features in real-time. This preliminary setup enables rapid detection during actual operation.
Solution Approach 2:
The patent replaces complex mechanical sensing systems with optical image recognition. Instead of using multiple sensors, encoders, or mechanical measurement devices on the operating arm, the system uses a simple camera to capture images and processes 2D image data to determine 3D pose. This substitution reduces hardware complexity and enables faster data acquisition while maintaining measurement capability.
3Measurement precision
If multiple pose identifications are used on the operating arm for accurate pose determination, then the pose measurement precision is improved, but the device complexity and identification difficulty increase
Solution Approach 1:
The patent divides the pose identification system into multiple independent marker balls distributed at different locations on the operating arm. Each marker ball is a simple, identical component with a unique coded pattern. By segmenting the measurement task across multiple identical simple units rather than using one complex identification system, the patent improves measurement precision through redundancy while keeping individual components simple and easy to manufacture.
Solution Approach 2:
The patent uses marker balls with different colored patterns or coded color arrangements as pose identifications. These color-coded markers provide high-contrast visual features that are easily distinguishable by the camera system. The color variations enable rapid automatic recognition and differentiation of multiple markers without requiring complex geometric features, simplifying the image processing while improving measurement accuracy.
Data Source
AI summary
The present application relates to the field of error detection technology. An error detection method and a robot system are provided. The error detection method includes: obtaining a target pose of an end of an operating arm; acquiring a positioning image; recognizing, in the positioning image, a plurality of pose identifications located on the end of the operating arm; recognizing, based on the plurality of pose identifications, an angle identification located on the end of the operating arm, the angle identification having a position association relationship with a first pose identification of the plurality of pose identifications; determining, based on the angle identification and the plurality of pose identifications, an actual pose of the end of the operating arm; and generating, in response to the target pose and the actual pose meeting an error detection condition, a control signal related to a fault.