3D Measurement Apparatus Epipolar Line Feature Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional three-dimensional measurement apparatuses face challenges in feature recognition performance due to overlapping codes caused by optical system blurs and the extraction of features perpendicular to projection and visual axes, leading to unmeasurable regions and degraded recognition accuracy.
Innovation Solution
A three-dimensional measurement apparatus that projects patterns with distinct feature patterns and epipolar lines, using a feature recognition unit to extract codes based on area ratios within defined regions, allowing for accurate feature recognition even when patterns overlap, and employing a three-dimensional shape calculation unit to associate measurement patterns and calculate the object's shape using triangulation principles.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If codes are extracted from captured feature portions that fluctuate perpendicularly to the plane formed by projection and visual axes, then codes can be associated with projected slits, but adjacent codes overlap due to optical system blur making specific codes unidentifiable
Solution Approach 1:
The invention changes the extraction direction from perpendicular to parallel with the plane formed by projection and visual axes. By extracting feature portions in the parallel direction (along epipolar lines), the patent avoids the overlap problem that occurs when extracting perpendicular to the plane, thereby maintaining code identifiability even with optical blur present
Solution Approach 2:
The invention converts the harmful effect of optical system blur into a manageable condition by changing the extraction approach. Instead of trying to overcome blur, the patent extracts features along epipolar lines where blur does not cause code overlap, effectively using the geometric constraints of the optical system to maintain measurement reliability
2Productivity
If features which fluctuate perpendicularly to the plane formed by projection and visual axes are extracted, then codes can be compared for association, but regions cannot be measured when these features overlap
Solution Approach 1:
The invention extracts feature portions parallel to the plane formed by projection and visual axes (along epipolar lines) rather than perpendicular to it. This dimensional change ensures that even when multiple projected slits are present, their corresponding feature portions do not overlap in the captured image, maintaining both measurement coverage and recognition performance
Solution Approach 2:
The invention segments the captured image into multiple epipolar lines, each corresponding to a specific projected slit. By processing each epipolar line independently and extracting features along these segmented lines, the patent prevents overlap between features from different slits while maintaining comprehensive measurement coverage
3Device complexity
If slits all having the same shape are projected, then the projection system is simple, but it is difficult to uniquely associate captured slits with projected slits unless measurement distance range is limited
Solution Approach 1:
The invention introduces epipolar lines as an intermediary geometric constraint to associate captured slits with projected slits. By defining epipolar lines based on the relative positions of the projection and capture apparatus, the patent creates a unique correspondence relationship without requiring complex coded patterns, thus maintaining simple projection hardware while achieving accurate slit association
Solution Approach 2:
The invention adds the dimension of epipolar line geometry to the association process. Instead of relying solely on shape differences or limited measurement ranges, the patent uses the geometric relationship between projection and capture axes to define epipolar lines, creating a new dimensional constraint that enables unique slit association across extended measurement ranges
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
Improves the probability of recognizing features even when they overlap due to optical system blurs, enhancing the overall feature recognition performance and enabling accurate three-dimensional shape calculation.
Implementation Method 1
an image of the object 108 is captured by the image capturing unit 101
Implementation Method 2
the image capturing unit 101 captures an image of the object 108
Data Source
Figure 1
Figure 2A~2B
Figure 3
AI summary
A three dimensional measurement apparatus comprising: projection means (102) for projecting a pattern including measurement and feature patterns onto an object; image capturing means (101) for capturing an image of the object onto which the pattern is projected; grouping means (106) for dividing the captured image into a plurality of regions using the measurement pattern in the pattern and a plurality of epipolar lines determined based on a positional relationship between the projection means and the image capturing means, thereby grouping predetermined regions among the plurality of divided regions; feature recognition means (106) for recognizing the feature pattern based on a difference in the feature pattern between the predetermined regions; and three-dimensional shape calculation means (107) for calculating a three-dimensional shape of the object based on the feature pattern.