Signed Distance Field for Local Geometry Deviation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for determining local deviations of an object's geometry from a target geometry are cumbersome and computationally intensive, requiring separate distance calculations between points on the target geometry and the object's surface.
Innovation Solution
The method utilizes a signed distance field (SDF) to directly provide distance information about local deviations, allowing the target geometry to be fitted into the distance field without needing to determine the position of the object, using techniques like least squares, minimum zone fit, or inscribed/circumscribing figures, and presenting deviations in false colors for easy identification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If separate distance calculations are performed between points on the target geometry and the object's surface, then measurement precision is improved, but computational load increases significantly
Solution Approach 1:
The patent pre-calculates and stores the distance transform of the object's distance field before comparison with target geometry. This preliminary computation allows rapid querying of distances during deviation analysis without performing repeated calculations, thus maintaining measurement precision while significantly reducing the computational load during the actual comparison phase
Solution Approach 2:
The patent creates a distance field representation (a digital copy) of the object's geometry that encodes distance information. This copied representation allows direct reading of distance values without performing geometric calculations on the original object model, enabling efficient deviation measurement while preserving accuracy
2Measurement precision
If the object's position must be determined before comparing with target geometry, then measurement accuracy is maintained, but the process complexity increases
Solution Approach 1:
The patent makes the distance field self-adaptive by having it automatically adjust to the object's actual position through gradient-based optimization. The distance field's gradient information guides the alignment process, allowing the system to self-correct positional deviations without requiring external position determination, thus maintaining measurement accuracy while eliminating the need for separate position determination steps
3Reliability
If multiple separate steps are used to determine position and then distance, then measurement reliability is improved, but the inspection time increases
Solution Approach 1:
The patent merges the position determination and distance measurement steps into a single integrated operation. By using the distance field's gradient information to simultaneously guide both positioning and measurement, the system achieves reliable deviation determination in one unified process rather than through separate sequential steps, thereby maintaining reliability while significantly reducing inspection time
Data Source
Figure 1
Figure 2
Figure 3a~3b
AI summary
The invention relates to a computer-implemented method for determining a local deviation of a geometry of an object from a target geometry of the object on the basis of a digital representation of the object, wherein the object representation comprises a plurality of image information items for the object. An image information item specifies a value of a measurand for the object at a defined position of the object. The method has the steps of determining the object representation, determining a distance field from the image information items of the object representation, wherein the distance field comprises a plurality of distance values, wherein a distance value for a specific point of the distance field specifies the shortest distance of the point from a closest material boundary (102) of the geometry of the object, determining the target geometry (108) of the object, recording the target geometry (108) of the object on the material boundaries of the geometry of the object specified by the distance values, and determining the local deviation of the geometry of the object from the target geometry (108) of the object at a test point on a material boundary predefined by the target geometry (108) by reading the distance value of the distance field at the position of the test point.