3D Feature Localization via Sensor Mapping for Noisy Welding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing depth-sensing camera systems struggle in environments with high noise levels and computational burdens, leading to inaccurate and inefficient performance in applications like automated welding, particularly in additive manufacturing.
Innovation Solution
A calibration process that maps an image sensor and a further sensor to a three-dimensional reference system, enabling accurate depth-sensing without requiring precise sensor positioning or emission of signals, and reducing computational load.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If depth-sensing camera systems use emission and reflection detection or stereo-vision with deep learning algorithms, then depth triangulation capability is improved, but computational burden increases and response time slows
Solution Approach 1:
The system performs calibration in advance to establish a mapping between sensor readings and three-dimensional positions. This pre-computed mapping enables real-time depth determination without requiring complex computational algorithms during actual operation, thus reducing computational burden while maintaining accuracy
Solution Approach 2:
The patent replaces complex computational algorithms (deep learning for image matching) with a simpler calibration-based approach. By substituting the mechanical/computational process of real-time image matching with a pre-established mapping relationship, the system achieves faster processing with reduced computational requirements
2Measurement precision
If depth-sensing camera systems use emission and reflection detection or stereo-vision, then depth triangulation capability is improved, but latency increases due to high computational burden
Solution Approach 1:
The calibration process establishes the mapping relationship between sensor data and three-dimensional positions beforehand. During real-time operation, the system simply queries this pre-computed mapping rather than performing complex calculations, dramatically reducing response latency while maintaining depth sensing accuracy
Solution Approach 2:
The system skips the time-consuming deep learning algorithm execution by using the pre-established calibration mapping. This allows the system to rapidly determine depth information by directly querying the mapping table rather than processing images through complex algorithms in real-time
3Measurement precision
If depth-sensing camera systems rely on signal emission and reflection detection, then depth triangulation is enabled, but performance deteriorates in environments with high noise from external light sources
Solution Approach 1:
The system uses a calibration mapping as an intermediary between the sensor readings and the three-dimensional position determination. This mapping, established under controlled calibration conditions, allows the system to interpret sensor data accurately even when external light conditions vary during operation, effectively filtering out noise interference
Solution Approach 2:
The calibration process captures sensor readings across multiple known three-dimensional positions and establishes a mapping that accounts for varying light conditions. By pre-characterizing the sensor response under different conditions during calibration, the system can compensate for noise from external light sources during actual operation
Data Source
AI summary
A method of calibrating a sensing system for determining a position of a feature of an object includes defining reference positions within a three-dimensional reference system; determining corresponding positions within an image frame of an image sensor and positions within a sensor frame of a further sensor; and generating a mapping linking the reference positions and the positions. A position of the feature of the object may be determined based on the mapping. There is also provided a method for monitoring a workpiece, the method comprising: obtaining shortwave infrared images; determining a corresponding image position of a feature of the workpiece or of an automated welding system using a machine-learned model; and determining a position of the feature.