3D Body Model Generation from 2D Images

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current three-dimensional modeling of the human body requires expensive or large sensors, limiting accessibility and practicality for generating dimensionally accurate models from two-dimensional images.

Innovation Solution

A method utilizing standard 2D cameras and machine learning techniques, such as convolutional neural networks (CNNs), to process two-dimensional body images from different views, segment body silhouettes, determine visibility indicators, and generate a dimensionally accurate three-dimensional model, including both visible and occluded body parts, with texture augmentation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of manufacture

If standard 2D cameras are used for three-dimensional body modeling, then device cost and complexity are reduced, but measurement precision and manufacturing precision deteriorate

Engineering Contradiction:
Improvedevice accessibilityVSAvoiddimensional accuracy
Core Design Contradiction:
Ease of manufactureVSManufacturing precision

Solution Approach 1:

The patent transforms 2D images from multiple viewing angles (front, back, left side, right side) into a 3D body model by adding the temporal dimension of sequential image capture and the spatial dimension of multi-angle perspective. This allows standard 2D cameras to generate dimensionally accurate body models without requiring specialized 3D sensors.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

Solution Approach 2:

The patent segments the body into multiple anatomical regions (head, torso, arms, legs) and processes each segment separately through silhouette extraction and parameter prediction. This segmentation approach enables accurate dimensional reconstruction of each body part from 2D images while maintaining overall body model precision.

Inventive Principle:
Principle #1Segmentation

2Loss of information

If multiple body images from different views are processed, then completeness of body model improves, but processing time increases

Engineering Contradiction:
Improvebody coverageVSAvoidprocessing duration
Core Design Contradiction:
Loss of informationVSLoss of time

Solution Approach 1:

The patent performs preliminary actions by pre-processing images to extract silhouettes and pre-identifying visible occluded landmarks before full 3D model generation. This preliminary processing of individual images prepares data structures that accelerate the subsequent integration phase, reducing overall processing time while maintaining complete body coverage.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent merges information from multiple 2D images taken from different views (front, back, left side, right side) into a unified 3D body model. By combining the silhouette data and landmark information from all views simultaneously, the system achieves complete body coverage including occluded regions without requiring sequential processing of each view.

Inventive Principle:
Principle #5Merging (Combining)

3Loss of information

If occluded body parts are included in the model, then model completeness improves, but measurement reliability becomes more challenging

Engineering Contradiction:
Improvevisibility coverageVSAvoidocclusion handling
Core Design Contradiction:
Loss of informationVSReliability

Solution Approach 1:

The patent creates virtual copies of visible body landmarks and uses them to infer the positions of occluded landmarks. By copying the spatial relationships and anatomical proportions from visible body parts, the system reliably reconstructs occluded regions without direct observation, maintaining model completeness while preserving measurement reliability through anatomical consistency.

Inventive Principle:
Principle #26Copying

Data Source

PatentUS11854146B1Three-dimensional body composition from two-dimensional images of a portion of a body
Publication Date: 2023.12.26 AMAZON TECH INC
  • US11854146B1 patent drawing
  • US11854146B1 patent drawing
  • US11854146B1 patent drawing

AI summary

Described are systems and methods directed to generation of a dimensionally accurate three-dimensional (“3D”) model of a body, such as a human body, based on two-dimensional (“2D”) images of at least a portion of that body. A user may use a 2D camera, such as a digital camera typically included in many of today's portable devices (e.g., cell phones, tablets, laptops, etc.) and obtain a series of 2D body images of at least a portion of their body from different views with respect to the camera. The 2D body images may then be used to generate a plurality of predicted body parameters corresponding to the body represented in the 2D body images. Those predicted body parameters may then be further processed to generate a dimensionally accurate 3D model or avatar of the body of the user.