Automated 3D Anatomy Generation via Template Interpolation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for generating animation-ready anatomies are labor-intensive and inefficient, particularly for applications requiring large numbers of animated models, such as visual effects and gaming, where manual creation scales poorly.
Innovation Solution
A method using a processor to interpolate a new anatomy's skeleton and skin mesh from selected template anatomies, based on shared poses and skinning weights, allowing for automated generation of animation-ready models by scanning an individual and accessing a database of template anatomies, with techniques like principal components analysis and weighted interpolation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If manual methods are used to create animation-ready anatomies, then each anatomy can be customized and animated, but the process is labor-intensive and does not scale well for large numbers of models
Solution Approach 1:
The patent uses template anatomies as source copies to generate target anatomies through automated interpolation. Multiple template anatomies with known animation data are combined to create new animation-ready models, eliminating manual creation for each model and enabling batch generation of large numbers of animated anatomies
Solution Approach 2:
The system performs self-service by automatically selecting appropriate template anatomies, computing interpolation weights based on skin mesh similarity, and generating animation data without human intervention. The automated workflow includes selecting templates, determining weights, and interpolating skeleton and skinning data to produce ready-to-use animated models
2Extent of automation
If template anatomies are selected from a collection, then the selection process can be automated based on skin mesh similarity, but additional processing steps are required to determine differences and select appropriate templates
Solution Approach 1:
The patent transforms the template selection problem into a parameter-based similarity comparison by computing skin mesh difference metrics. Templates are selected based on quantitative parameters (skin mesh similarity scores) rather than manual evaluation, enabling automated selection while managing complexity through systematic parameter comparison and weighting
3Manufacturing precision
If interpolation is performed from multiple template anatomies, then the quality and accuracy of the generated anatomy can be improved, but the computational complexity and processing time increase
Solution Approach 1:
The patent applies local quality by assigning different interpolation weights to different template anatomies based on their skin mesh similarity to the target. Templates with higher similarity receive greater weights, allowing the system to prioritize more relevant templates and improve accuracy while managing computational complexity through selective weighting rather than uniform processing of all templates
Data Source
AI summary
Methods, systems, and techniques for generating a new, animation-ready anatomy. A skin mesh of the new anatomy is obtained, such as by performing a 3D depth scan of a subject. Selected template anatomies are also obtained, with each of those template anatomies having a skin mesh that corresponds with the new anatomy's skin mesh. The skin meshes of the new and selected template anatomies share a first pose. Each of the selected template anatomies also has a skeleton for the first pose and skinning weights, and the skin mesh in at least one additional pose that is different from the first pose and any other additional poses. The method then involves using a processor to interpolate, from the at least one of the skeleton and skinning weights of the selected template anatomies and the first and at least one additional pose of the selected template anatomies, the new anatomy.