3D Body Geometry Archiving via PCA Vectorization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems for storing and retrieving 3D geometry data of the human body require large datasets, leading to high storage demands and inefficiencies due to noise, artifacts, and varying levels of detail needed for different applications.
Innovation Solution
A method and system that reduce data storage by organizing and analyzing 3D geometry information using vectorization and covariance analysis to identify scalable characteristic components and scale factors, allowing for approximation of body geometry with reduced data, tailored to specific applications and purposes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If large datasets are used to store 3D geometry information, then measurement precision is improved, but storage requirements increase
Solution Approach 1:
The patent extracts and removes noise and artifacts from the original 3D scan data, keeping only the meaningful geometric information. This extraction process eliminates irrelevant data that contributes to storage requirements without providing useful measurement information, thereby reducing the quantity of data to be stored while preserving measurement precision.
Solution Approach 2:
The patent transforms the 3D geometry data from a complete detailed model into a set of statistical parameters including mean shape, principal component analysis (PCA) coefficients, and covariance matrices. This parameter transformation dramatically reduces the data dimensionality while retaining the essential geometric characteristics needed for accurate representation and measurement.
2Loss of information
If complete 3D geometry data is stored, then information completeness is improved, but storage requirements increase
Solution Approach 1:
The patent performs preliminary processing of the 3D scan data by computing statistical parameters (mean shape, PCA modes, covariance) in advance. This preliminary action creates a compressed statistical representation that captures the essential variability of the geometry, allowing complete information to be represented in a much smaller data structure than the original detailed mesh or point cloud.
3Measurement precision
If detailed 3D geometry information is stored, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The patent replaces complex geometric processing operations with statistical computations. Instead of manipulating detailed geometric models directly, the system uses statistical parameters and linear algebra operations (PCA, covariance analysis) to achieve the same measurement objectives with simpler, more efficient computational methods.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method and system for archiving 3D geometry information for a plurality of subjects, and on a particular part of the body (which may also be the whole body) is described. The method and system allow an organizing and an analyzing step, resulting in scalable characteristic components for the particular part of the body, and an approximation of the geometry of that particular part of the body for a specific subject by a combination of a subset of those characteristic components with corresponding scale factors, which are then stored. The method comprises the steps of : ° organizing for each of said plurality of subjects the 3D geometry information according to a predetermined standard, thus obtaining organized 3D geometry information; ° analyzing the organized 3D geometry information of said plurality of subjects, resulting in averaged 3D geometry information of said particular part of the body; and - scalable characteristic components of said particular part of the body; such that for each of said plurality of subjects said organized 3D geometry information can be approximated by a combination of said scalable characteristic components relative to said averaged 3D geometry information; ° comparing, for at least one out of said plurality of subjects, said organized 3D geometry information with said averaged 3D geometry information and determining a subset of said scalable characteristic components and corresponding scale factors, for approximating to a given accuracy said organized 3D geometry information relative to said averaged 3D geometry information by a combination of said scalable characteristic components out of said subset, each characteristic component scaled with its corresponding scale factor; ° storing, for at least one specific subject out of said plurality of subjects, said approximated organized 3D geometry information by storing - said averaged 3D geometry information; - said subset of said scalable characteristic components and said corresponding scale factors for said specific subject.