Iterative Mesh Generation for 3D Body Shape Modeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for generating three-dimensional representations of human bodies are inefficient in capturing accurate shapes, particularly in commercial applications like virtual fitting, due to issues with mesh fitting and resolution, leading to over-stretched quads and limited access to individual body parts.
Innovation Solution
The method involves subdividing patches corresponding to faces of blocks formed from cubes, iteratively fitting meshes to point cloud data, and generating a statistical model that represents the three-dimensional shape of a class of subjects, such as humans, using a system with a user interface, processor, and point cloud capture device to output a high-resolution mesh.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If a mesh is fitted to point cloud data using conventional algorithms, then the fitting process achieves computational efficiency, but the resulting mesh contains over-stretched quads and fails to accurately represent individual body parts
Solution Approach 1:
The body model is segmented into multiple patches corresponding to different body parts (torso, limbs, head). Each patch is independently parameterized and fitted to the corresponding region of the point cloud data, allowing for accurate representation of individual body parts while maintaining overall mesh coherence.
Solution Approach 2:
Different regions of the mesh are assigned different levels of detail and parameterization density based on their importance and geometric complexity. Critical body parts receive higher resolution parameterization, while less critical regions use coarser parameterization, optimizing both accuracy and computational efficiency.
2Measurement precision
If the mesh resolution is increased to capture accurate body shapes, then the representation precision improves, but the computational complexity and processing time increase significantly
Solution Approach 1:
A coarse mesh is generated first as a preliminary structure that captures the overall body shape. This coarse mesh is then systematically refined through iterative subdivision and local optimization, allowing the algorithm to converge to high precision without the computational burden of starting with a fine mesh.
Solution Approach 2:
The mesh structure is made dynamically adaptive during the fitting process. The parameterization density and mesh resolution are adjusted iteratively based on the local geometric features and fitting residuals, concentrating computational effort where it is most needed while maintaining efficiency in well-fitted regions.
3Ease of operation
If conventional fitting algorithms are used to generate three-dimensional models, then the process is computationally efficient, but the resulting models have limited access to individual body parts and poor regularity
Solution Approach 1:
The parameterized patch structure provides a universal framework that serves multiple functions: it enables accurate geometric representation, facilitates easy access to individual body parts through parameter space navigation, supports various rendering and analysis operations, and maintains regularity across different body regions through consistent parameterization schemes.
Data Source
AI summary
In one embodiment a method of generating a three dimensional model of a subject is disclosed. The method comprises receiving point cloud data for a subject; receiving user inputs indicating a plurality of cubes, and a grouping of the cubes into a plurality of blocks to form a representation of a class of subjects; generating a first mesh comprising a plurality of quadrilaterals by subdividing patches corresponding to faces of blocks of the plurality of blocks; fitting the first mesh to the point cloud data to generate a fitted mesh; iteratively generating further meshes, each comprising a plurality of quadrilaterals by subdividing patches of the fitted mesh from the previous iteration and fitting the further mesh to the point cloud data; and outputting as the three dimensional model of the subject the iteratively generated fitted mesh.