Digital Human Model Vision Constraint for Posture Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing digital human models (DHMs) fail to accurately account for target size, orientation, and precision level when predicting human postures, leading to unrealistic simulations due to insufficient vision constraints.
Innovation Solution
A novel vision constraint method that considers target size, orientation, and precision level, using a data engine to obtain posture information, a processing engine to generate a vision measure, and a posturing engine to update the DHM posture, incorporating parameters like head-target distance and facet orientations to improve simulation accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing vision constraints are used in DHM, then computation is simpler, but accuracy of posture simulation is insufficient because target size, orientation, and precision level are not accounted for
Solution Approach 1:
The patent introduces a comprehensive vision constraint model that incorporates multiple parameters including target size, target orientation, and precision level, transforming the simple vision constraint into a multi-parameter constraint system. This allows the DHM to accurately simulate postures by considering these additional parameters that affect visual task performance.
2Measurement precision
If a comprehensive vision constraint model accounting for target size, orientation, and precision level is implemented, then accuracy of posture simulation is improved, but computational complexity increases
Solution Approach 1:
The patent applies local quality by implementing vision constraints specifically at the head/target interaction level rather than globally across the entire DHM system. The vision constraint model focuses computational resources on the critical visual task parameters (target size, orientation, precision level) while leaving other body segments to be determined by standard inverse kinematics, thus balancing accuracy with computational efficiency.
3Reliability
If vision constraints are added to DHM, then realism of simulated posture is improved, but number of constraints and variables increases
Solution Approach 1:
The patent creates a universal vision constraint model that can handle multiple types of visual tasks by incorporating target size, orientation, and precision level parameters. This multi-functional constraint system can adapt to different task requirements without requiring separate constraint models for each task type, thus improving realism while managing the number of constraints through parameterization rather than multiplication of separate constraint equations.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present disclosure is directed to a method and corresponding system that improves accuracy of a computer simulation of an original posture of a digital human model (DHM) relative to a target object. The method and system may obtain information associated with the original DHM posture. The obtained DHM posture information may include a position of a head of the DHM. The method and system may obtain information associated with the target object. The obtained target object information may include a size of the target object and an orientation of the target object. The method and system method may obtain a distance from the head of the DHM to the target object. In some embodiments, the system and method may generate a measure of vision (i.e., vision measure) of the DHM of the target object that the DHM is visually targeting. The system and method may generate the measure of vision based on one or more parameters which may include any of the obtained DHM posture information, the obtained target object information, and the obtained head-target (HT) distance. Based on the measure of vision, the system and method may generate a constraint of vision (i.e., vision constraint) of the DHM to the target object. Based on the vision constraint, the system and method may generate an updated DHM posture.