3D Object Tracking via Polygonal Plane Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing 3D model-based tracking methods fail to achieve precise object tracking due to incorrect calculation of the 3D rigid body transform matrix when points are not correctly matched to the object's edge, leading to inaccurate object positioning.
Innovation Solution
Divide the object into a plurality of polygonal planes and track using texture data and geometric data between planes, generating virtual planes when the tracking failure rate is below a designated level and stopping tracking when it exceeds, to ensure accurate object positioning.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If 3D model-based tracking is performed using edge-based tracking of a single 3D model, then the tracking process can be implemented, but the tracking precision deteriorates when points are not correctly matched to the object's edge
Solution Approach 1:
The patent divides the single 3D model into multiple polygonal planes (e.g., front surface, side surfaces, top surface) and tracks each plane independently. This segmentation allows the system to handle complex objects with varying geometries more reliably, as each plane can be tracked using its specific texture and geometric characteristics, reducing the impact of incorrect point matching on overall tracking precision.
Solution Approach 2:
The patent applies different tracking methods to different planes based on their specific characteristics. Each plane is tracked using its unique texture data and geometric constraints, allowing localized optimization of tracking precision for each surface region rather than applying a uniform tracking approach to the entire model.
2Manufacturing precision
If a single 3D model is used for tracking, then the model structure is simple, but the tracking accuracy deteriorates due to incorrect calculation of 3D rigid body transform matrix
Solution Approach 1:
The patent segments the 3D model into multiple polygonal planes, each with its own tracking parameters and transform matrices. This segmentation enables more accurate tracking by allowing independent optimization of each plane's position and orientation, while the overall model structure remains manageable through hierarchical organization of planes.
Solution Approach 2:
The patent transitions from tracking a single monolithic 3D model to tracking multiple 2D planes that compose the 3D object. This dimensional decomposition allows for more precise texture matching and geometric constraint application on each plane, improving overall tracking accuracy without requiring complex computations on the entire 3D model.
3Reliability
If edge-based tracking is performed on the 3D model, then the tracking process can be executed, but the reliability deteriorates when points do not meet the edge of the 3D model correctly
Solution Approach 1:
The patent divides the object into multiple polygonal planes and tracks each plane independently using its specific texture and geometric data. This segmentation ensures that point matching is performed locally on each plane's edge, improving reliability by reducing the propagation of errors from incorrect point matching to the overall tracking result.
Solution Approach 2:
The patent incorporates feedback mechanisms to monitor tracking quality and generate virtual planes when tracking failures are detected. This feedback system allows the system to adapt to incorrect point matching by generating corrective virtual planes, thereby maintaining positioning precision even when initial edge-based tracking fails.
4Duration of action of stationary object
If virtual planes are generated when tracking failure rate is below designated level, then tracking continuity is improved, but the system complexity increases
Solution Approach 1:
The patent prepares virtual planes in advance as a backup mechanism. When tracking failures are detected, these pre-prepared virtual planes can be quickly generated and applied to maintain tracking continuity without requiring complex real-time computations or system reconfiguration, thus reducing the actual system complexity burden.
Solution Approach 2:
The patent introduces virtual planes as intermediary elements that mediate between the original 3D model and the tracking system. These virtual planes serve as placeholders or corrective elements when tracking failures occur, allowing the system to maintain continuity without directly modifying the original model structure, thereby managing system complexity effectively.
Data Source
AI summary
A 3D object tracking method and apparatus in which a model of an object to be tracked is divided into a plurality of polygonal planes and the object is tracked using texture data of the respective planes and geometric data between the respective planes to enable more precise tracking. The 3D object tracking method includes modeling the object to be tracked to generate a plurality of planes, and tracking the plurality of planes, respectively. The modeling of the object includes selecting points from among the plurality of planes, respectively, and calculating projective invariants using the selected points.