3D Marker Construction for AR Pose Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for detecting the three-dimensional pose of objects in augmented reality and medical contexts face challenges such as marker visibility issues due to lighting conditions and cumbersome setup requirements, particularly for objects like surgical tools.
Innovation Solution
A method for constructing a three-dimensional marker from a precursor element by determining two-dimensional representations and constructing sections with varying heights based on these representations, allowing for optimized pose determination and adaptation to object dimensions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If two-dimensional markers are used for object pose detection, then the implementation is simple, but the marker may not be visible when the object is moved or under poor lighting conditions
Solution Approach 1:
The patent transitions from two-dimensional markers to three-dimensional markers with varying heights. The marker includes multiple columns with different heights that create a three-dimensional structure, allowing depth cameras to detect the marker from various angles and orientations, thereby solving the visibility problem while maintaining implementation simplicity.
Solution Approach 2:
The patent changes the physical parameter of the marker from flat two-dimensional to three-dimensional by varying the heights of columns. This parameter change enables the marker to be detected in different lighting conditions and object orientations, improving reliability while keeping the marker structure relatively simple.
2Reliability
If depth cameras and three-dimensional point cloud analysis are used to avoid lighting problems, then lighting robustness is improved, but the setup becomes cumbersome requiring object placement at the center of cube arrangements
Solution Approach 1:
The patent uses a single three-dimensional marker with height variations instead of multiple cubes arranged in specific configurations. The three-dimensional structure provides sufficient depth information for reliable detection without requiring complex spatial arrangements, simplifying the setup process while maintaining lighting robustness.
Solution Approach 2:
The three-dimensional marker serves multiple functions: it provides lighting-robust detection, enables pose determination from various angles, and eliminates the need for complex cube arrangements. This universal marker design simplifies the overall system setup while maintaining reliability across different conditions.
3Measurement precision
If three-dimensional markers with height variations are constructed from two-dimensional representations, then pose determination accuracy is improved, but the manufacturing process becomes more complex
Solution Approach 1:
The patent divides the marker into multiple columns with different heights, where each column's height is determined by values from a two-dimensional representation. This segmentation approach allows the complex three-dimensional structure to be constructed from simpler two-dimensional data, improving pose accuracy while managing manufacturing complexity through systematic division.
Solution Approach 2:
The patent uses a two-dimensional representation (matrix of values) as a template to generate the three-dimensional marker structure. The height of each column is copied from corresponding values in the two-dimensional representation, allowing accurate pose determination while simplifying the manufacturing process by using a digital template that can be easily reproduced.
Data Source
AI summary
A method for constructing a 3D label from a precursor element and the 3D label are provided, the precursor element including at least one surface, the method including: determining at least one 2D representation associated with the at least one surface of the precursor element, the 2D representation comprising a plurality of values associated with parts of the surface of the precursor element, respectively, and constructing a corresponding section of the 3D label from the surface by attaching a structural feature of an element of the respective section, defining said element from the surface according to a value of the 2D representation associated with the respective part of the surface.