3D Body Reconstruction from Depth Images via Parametric Model Fitting

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for generating three-dimensional representations of human bodies from depth images are inefficient and require manual marker placement, limiting their usability and deployment in applications such as medical visualization and virtual fitting.

Innovation Solution

A system that uses parametric models to fit a three-dimensional representation of a human body by minimizing an energy function, which includes a distance term that prioritizes points inside the volume defined by the depth image, and incorporates gravity functions to normalize and reduce the resolution of the depth image, allowing for automatic generation of 3D data without manual markers.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual marker placement is used for 3D body representation, then measurement precision can be improved, but ease of operation and time consumption deteriorate

Engineering Contradiction:
Improvebody shape estimation accuracyVSAvoidoperator intervention requirement
Core Design Contradiction:
Measurement precisionVSEase of operation

Solution Approach 1:

The system automatically detects body landmarks and performs 3D reconstruction without requiring manual marker placement by operators. The algorithm autonomously identifies key body points from depth images and computes the 3D representation, making the system self-sufficient and eliminating manual intervention while maintaining accuracy

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent replaces the mechanical process of manual marker placement with an automated computational system that uses depth image processing and algorithms to detect body landmarks and generate 3D models, substituting human manual operations with automated image analysis

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Manufacturing precision

If complex processing algorithms are used for accurate 3D reconstruction, then manufacturing precision improves, but productivity deteriorates

Engineering Contradiction:
Improve3D representation accuracyVSAvoidgeneration speed
Core Design Contradiction:
Manufacturing precisionVSProductivity

Solution Approach 1:

The 3D reconstruction process is divided into distinct segments: depth image acquisition, body landmark detection, parametric model fitting, and 3D model generation. This segmentation allows each component to be optimized independently, maintaining high accuracy while improving overall processing efficiency through modular computation

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system performs preliminary processing on depth images including noise filtering, normalization, and key feature detection before the main 3D reconstruction. This preliminary action prepares the data in advance, reducing the computational burden during the actual reconstruction phase and accelerating the overall process while preserving accuracy

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP2819098B1Methods and systems for generating a three dimentional representation of a subject
Publication Date: 2018.09.26 KK TOSHIBA
  • EP2819098B1 patent drawingFigure 1
  • EP2819098B1 patent drawingFigure 2~3
  • EP2819098B1 patent drawingFigure 4a~4d

AI summary

According to one embodiment, a method of generating a three dimensional representation of a subject from a depth image, comprises comparing a depth image of the subject with a plurality of representative images, wherein each representative image is associated with a respective parametric model of a subject; identifying a representative image of the plurality of representative images as a closest representative image to the depth image of the subject; selecting the parametric model associated with the closest representative image to the depth image; and generating a three dimensional representation of the subject by fitting the selected parametric model to the depth image of the subject.