3D Coordinate Determination via Multi-Camera Photogrammetry
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for determining 3D coordinates of objects using fringe projection systems are not sufficiently accurate, especially for larger objects with limited surface structure, and require extensive setup and calibration, often restricting robot movement and object accessibility.
Innovation Solution
A method employing a projector, camera, and multiple reference cameras with fixed orientations, combined with an industrial robot for precise positioning and calibration, using photogrammetry to determine the positions and orientations of reference marks and the measuring system, allowing for global registration and correction of robot positions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single fringe projection system is used to determine 3D coordinates, then the measurement process is simple, but the accuracy is insufficient for larger objects with little surface structure
Solution Approach 1:
The patent divides the measurement system into multiple fringe projection systems positioned at different recording positions, each capturing different portions of the object. The measurements from these segmented systems are then combined through global registration to achieve comprehensive and accurate 3D coordinate determination for large objects.
Solution Approach 2:
The patent introduces reference marks as intermediary elements that are captured by both the fringe projection system and a separate measuring system (such as a laser scanner). These reference marks serve as common reference points that enable accurate transformation and registration of measurements from different systems and positions into a unified coordinate system.
2Area of stationary object
If multiple recordings are made at different positions to capture complete objects, then the object coverage is improved, but the global registration accuracy deteriorates due to cumulative errors
Solution Approach 1:
Reference marks serve as stable intermediary reference points that are visible in multiple recordings from different positions. By transforming all recordings to a common coordinate system based on these shared reference marks, the patent eliminates cumulative registration errors and achieves high-precision global registration across the entire object surface.
Solution Approach 2:
The patent uses reference marks with distinct visual characteristics (such as contrasting colors or patterns) that make them easily distinguishable from the object surface. This allows the measuring system to reliably detect and use these marks for coordinate transformation even when the object itself has little surface structure or uniform appearance.
3Ease of manufacture
If robot position information is used for global registration, then the setup effort is reduced, but the registration accuracy is insufficient
Solution Approach 1:
The patent uses measured reference marks to provide feedback on the actual positions of the fringe projection system relative to the object. This measured information is used to correct and refine the robot position information, transforming approximate robot-based registration into accurate measurement-based registration without requiring manual setup of complex reference systems.
Solution Approach 2:
The patent replaces reliance on mechanical robot position information with optical measurement-based positioning using reference marks. Instead of using the robot's mechanical coordinates (which have limited precision), the system uses optically measured reference points to determine positions, substituting a mechanical approach with an optical measurement approach for higher accuracy.
4Measurement precision
If reference backdrops are placed around the object for measurement, then the global registration is enabled, but the object accessibility and robot movement are restricted
Solution Approach 1:
The patent extracts the reference marks from being part of a fixed backdrop structure and places them directly on or near the object surface. This allows the reference marks to be captured without requiring large backdrop structures that would restrict robot movement or block access to the object, while still providing the necessary reference points for global registration.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach enhances the accuracy of 3D coordinate determination and robot calibration, enabling efficient global registration and improved accessibility for objects of varying sizes, even when reference marks restrict movement, by using multiple reference cameras and photogrammetric bundle block adjustments.
Implementation Method 1
a projector for projecting a fringe pattern onto the object and a camera for recording the fringe pattern reflected back from the object
Implementation Method 2
The recording is evaluated by an evaluation system that includes a computer... The measuring marks are first measured. This is preferably done using the photogrammetry method.
Data Source
Figure 1
AI summary
The device has a projector (10) for projecting a pattern i.e. stripe pattern, of an object (1), and a camera (11) connected with the projector for receiving the object. Multiple reference cameras (16-18) receive multiple encoded reference marks (6, 24) of a reference mark field (25) and are connected with the projector and the camera. The reference cameras are arranged in a camera module (15). An evaluating device i.e. personal computer, determines position and orientation of the projector, the camera, the reference cameras and a measuring system (23) that calibrates an industrial robot. Independent claims are also included for the following: (1) a method for determining three-dimensional (3D)-co-ordinates of an object (2) a method for calibrating an industrial robot.