Markerless Motion Analysis with Enhanced 3D Angular Kinematics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing markerless motion capture systems struggle to accurately measure three-dimensional rotational motions due to a limited set of spatial coordinates, leading to inaccurate human pose and spatial landmark identification, especially in dynamic activities with environmental variations.
Innovation Solution
Enhance 3D angular kinematic data using model equations and probabilistic mappings, combining musculoskeletal body models with anatomical constraints, and employing supervised or machine learning techniques to estimate 3D spatial orientations from under-constrained body-fixed reference points.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If markerless motion capture is used to avoid wearing markers and placing cameras accurately, then ease of operation is improved, but measurement precision of 3D angular kinematics deteriorates
Solution Approach 1:
The patent introduces deep learning models as intermediary components that process markerless motion capture data and generate enhanced 3D angular kinematic data. These models act as mediators between the limited input data from markerless systems and the required accurate output measurements, translating imprecise camera-based observations into reliable kinematic parameters through learned transformations.
Solution Approach 2:
The patent transforms the approach by changing from direct measurement of 3D angular kinematics to measuring 2D image parameters and using deep learning models to infer the 3D kinematics. This parameter transformation allows the system to work with the limited data from markerless capture while achieving accurate 3D motion analysis through computational enhancement.
2Ease of operation
If typical video or imaging analysis systems are used to capture golf swings, then ease of operation is improved, but measurement precision of human pose and spatial landmarks deteriorates
Solution Approach 1:
The patent replaces traditional mechanical marker-based measurement systems with a computational approach using deep learning models. Instead of relying on physical markers and complex camera calibration mechanisms, the system uses neural networks to directly infer 3D angular kinematics from standard video images, substituting mechanical measurement infrastructure with intelligent computation.
Solution Approach 2:
The system changes the measurement approach by using 2D image parameters from standard cameras and transforming them into 3D angular kinematic data through deep learning models. This parameter transformation enables the use of simple video capture equipment while achieving precise motion analysis that would otherwise require complex marker-based systems.
3Measurement precision
If marker-based motion analysis systems are used to achieve accurate 3D rotational motions, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The patent extracts the complex computational processing requirements from the physical measurement system. By removing markers and camera calibration complexity, the system isolates the motion capture to simple video recording, then handles all the complexity of 3D reconstruction and kinematic calculation through deep learning models in the software domain.
Solution Approach 2:
The patent substitutes mechanical marker-based measurement infrastructure with computational deep learning models. This replacement eliminates the need for physical markers, specialized cameras, and complex calibration procedures, achieving accurate 3D angular kinematics measurement through software-based inference rather than hardware complexity.
Data Source
AI summary
Systems, methods, and computer-readable storage devices are disclosed for improving markerless motion analysis. One method including: receiving position data of joint centers of a body in motion captured by at least one camera; enhancing, using model equations, three-dimensional (3D) angular kinematic data of the position data of the joint centers of the body, wherein the enhanced 3D angular kinematic data includes increased measurement accuracy of the position data of the joint centers of the body; and providing the enhanced 3d angular kinematic data for display to evaluate motion performance.