Presenter: Dr Gunther PaulCo-Author: Nathan Daniell

ErgoLab: Mawson InstituteUniversity of South Australia

3D Anthropometry – From the Ergonomist’s Perspective

University of South Australia

How does 3D scanning work?

A laser stripe is projected onto the body with the reflection captured by cameras

Body represented by series of XYZ data points, often referred to as “point cloud” data

Coordinates calculated by triangulation method

Camera (height known)

Camera (height known)

Laser (height known)β

University of South Australia

Scanner Features

Feature VITUS XXLHamamatsu

BodyLine Scanner

Cyberware WBX

Colour Scanner

TC2 NX16

Light source Laser Laser Laser White light

Cameras 8 4 4 16

Volume (cm)

H x W x D

210 x 120 x 100 200 x 100 x 60 200 x 50 x 130 200 x 120 x

60

Scan time (s) 12 5/10 < 20 8

V-Resolution

pitch (mm)

~3 2.5/5 2 2

Colour orTexture Texture Luminance Colour No

Software Anthroscan Body Line Manager Digisize TC2 Msr

Software

2 mm

2 mm

~0.5 mm

Accuracy: 4.03 mm

Measurement Extraction

Data Analysis

Real Bodies

Physical Measurements

Anthropometrist

Extract 1 and 2D Measurements

Manual extraction

Clean Scan

Remove artefacts

Point Cloud

Scan of human

Landmarked Bodies

Anthropometrist

Filled Scan

File conversion

Extract 3D measurements

Manual extraction

Automatic measurement

extraction

Scan of human, automatic extraction

University of South Australia

Trips and fallsThe unknowns: Limited access to scanner calibration data Validity of gap filling function Few studies validating segmental 3D measurements

Other limitations: High cost Automatic landmarker identification File formats

University of South Australia

Quality assessment

Automatic Landmark Function: TC2 (2002) compared chest and hip circumference. The bias

recorded was -47.0 to -27.9 mm. Mckinnon & Istook (2001) scanned a mannequin and extracted 10

measurements. The bias recorded was ± 19.3 mm.

Segmental volume measurements: Norton et al. (2002) compared segmental volume of the leg.

Average difference was 0.61 %. Wells et al. (2006) compared 5 segments using a mannequin.

Average differences ranged from 0-5%.

Segmental measurements have not been validated using an object that accurately represents a human

University of South Australia

ISO/FDIS 20685

three-dimensional3-Dpertaining to the use of three orthogonal scales on which the three coordinates, x, y and z, can be measured to give the precise position of any relevant anatomical point in the considered space

3-D body scannerhardware and software system that creates digital data representing a human form, or parts thereof, in three dimensions

3-D scanning methodologies for internationally compatible anthropometric databases

University of South Australia

Lack of standardisation

There are currently no standardized methods for using 3-D point clouds in the design process.

As a result, many users extract one-dimensional (1-D) data from 3-D point clouds.

This International Standard concerns the application of 3-D scanners to the collection of one-dimensional anthropometric data for use in design.

It does not apply to instruments that measure the location and/or motion of individual landmarks. (as defined in ISO 7250-1:2008: Basic human body measurements for technological design — Part 1: Body measurement definitions and landmarks)

University of South Australia

Landmarks should be

marked on the skin, and then identified with dots or other techniques that

can be seen on the displayed image and distinguished using the available software

Bilateral landmarks should be marked on both sides of the body.

If landmarks (see Clause 3) are to be marked before scanning, a minimal list would be the following: (21 landmarks)

ISO/FDIS 20685University of South Australia

ISO/FDIS 20685 There are a number of different fundamental technologies that

underlie commercially available systems.

These include stereophotogrammetry, ultrasound and light (laser light, white light and infrared).

Further, the software that is available to process data from the scan varies in its methods.

Additionally, software to extract dimensions similar to traditional dimensions varies markedly in features and capabilities.

As a result of differences in fundamental technology, hardware and software, extracted measurements from several different systems can be markedly different for the same individual.

[The required accuracy is 1-9 mm, depending on the measurement.]

University of South Australia

Anthropometric Design

Human boundary measures in space

Functional measures (Reach, Vision etc.)

Anthropometric parameters (to be considered in physiology, biomechanics etc.)

Any cross section between two volumetric objects in 3D space

forms a plane Ergonomic Design is defined in planes: inside dimensions,

outside dimensions, functional space 3D for visualization only

2D

University of South Australia

Anthropometric DesignUniversity of South Australia

Anthropometric DesignUniversity of South Australia

Anthropometric DesignUniversity of South Australia

Female Male comparison All roads combined

Female Sample size 42

Male Sample size 34

FemaleMale

2468101214161820222426283032

Anthropometric Design

Traditional approach

measure 1D Anthropometry option to “enrich” the data with 2D curves or splines create 2D/3D manikins

“3D” approach

measure 3D Anthropometry “reduce” the data with multiple filters to 2D curves or splines “reduce” the data with multiple filters to 1D anthropometric

dimensions

University of South Australia

True BenefitUniversity of South Australia

Digitised data

Ability to store and access at a later date

Less invasive

Reduces time required by subject

Allows to model real persons into 3D manikins using point cloud

SummaryUniversity of South Australia

Quality issues not solved

Precision not conform with ISO 20685

Combining data from 1D and 3D scans remains difficult

Economic benefits

Not required for ergonomic, anthropometric design

Main advantage is visualization