Human Motion Simulation Using Relative Joint Angle Changes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current human movement simulation tools are cumbersome, requiring skilled operators and empirical data that is limited by joint angle definitions and landmark data, leading to unrealistic simulations and difficulty in modeling movements from arbitrary start postures.
Innovation Solution
A computerized method that stores and compares data sets of human movements to simulate desired movements based on start and end points, using relative joint angle changes to adapt to any posture and kinematic configuration, eliminating the need for mathematical optimization and allowing for realistic simulation of human motion.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If motion capture technology with joint angle data is used to model human movement, then movement data can be captured empirically, but the data cannot be applied to arbitrary skeletal configurations because angle definitions are dependent on the skeletal configuration
Solution Approach 1:
The patent transforms the representation of movement data by changing from joint-angle-based parameters to landmark-based parameters. This parameter transformation allows the same movement data to be applied across different skeletal configurations without requiring redefinition of angles for each configuration, thereby resolving the contradiction between adaptability and complexity.
2Measurement precision
If mathematical optimization methods are used to fit kinematic skeleton to landmark data, then constraint solutions can be obtained, but the methods are slow and inconsistent
Solution Approach 1:
The patent replaces the mathematical optimization mechanism with a direct geometric computation mechanism. Instead of using iterative optimization algorithms to fit the skeleton to landmarks, the system uses closed-form geometric solutions that compute the fit directly, eliminating the time-consuming iterative process while maintaining precision.
3Reliability
If empirical data from lab observations are used for simulation, then real human movement patterns can be captured, but the data reflects only specific experimental conditions and cannot generalize to arbitrary tasks and loading conditions
Solution Approach 1:
The patent creates a universal movement representation system using landmark data that can be applied across multiple tasks and conditions. By representing movement as transformations of landmark positions rather than task-specific joint angles, the system achieves multi-functionality where the same data structure serves various simulation scenarios including different tasks, loading conditions, and skeletal configurations.
4Ease of operation
If key frame locations with simple posture transition interpolators are used, then computational simplicity is achieved, but the resulting motion looks robotic and unrealistic
Solution Approach 1:
The patent copies the temporal patterns and relative motion relationships from real human movement data and applies them to the simulated motion. By preserving the timing, duration, and relative progression of movements from empirical observations while using simplified landmark-based representation, the system generates motion that is both computationally simple and realistically human-like.
Data Source
AI summary
According to one embodiment of the invention, a computerized method for simulating human movement includes storing a plurality of sets of data, in which each set of data is indicative of a measured movement of a first human, receiving a start point and an end point for a desired movement of a second human, and comparing the desired movement to the stored sets of data. The method further includes selecting, based on the comparison, a stored set of data that is representative of the desired movement and simulating the desired movement based on the start point, the end point, and the relative change in position of a first joint associated with the selected set of data from an empirical start point to an empirical end point.