Cross-platform motion capture data sharing

A case study: Vicon and BTS

Dr.Isabelle Cossette & Phil Dixon

Presentation outline

• Goals: What can be learned?

• Background: Analysis of Respiration viamotion capture techniques.

• Introduction: The BTS Smart motion CaptureSystem.

• Smart file types: A look at .tdf files.

• File Sharing (part I): Vicon to Smart.

• File Sharing (Part II):Smart to Vicon.

• Final Remarks

Goals

• Uncover common difficulties in file sharing

between systems.

• Find possible common solutions.

• Incite discussion.

Background: Analysis of Respiration

• Scientific approach to breathing in music

performance.

• Important respiratory variables (chest wall volumes,

mouth pressure, flow rates, etc...) can be anaylsed

using motion capture techniques and optoelectronic

plethysmography (OEP).

• Main Goals of Dr. Cossette’s research:

-Evaluating the validity of popular beliefs.

-Helping performers.

-Accelerating and improving teaching.

Jugular Notch

Angle of Louis

Nipple

Xiphoid

Pulmonary

Rib Cage

Lower Costal Margin

Abdominal

Rib Cage

Ombelicus

Ant Sup Iliac Crest

Abdomen

OEP

•OEP: Measures variation in chest wall volumes bytracking the position in space of a series of markersplaced on anatomical landmarks along the chest wall.

Background: Analysis of Respiration (con’t)

The BTS Smart motion Capture System.

• BTS created in 1986 at polytechnic of Milan.

• BTS system:

-Calculates important respiratory parameters

(vital capacity, respiratory frequency,

inspiration/expiration flux, etc…)

-Contains 9 camera system

-Contains 3 separate software modules

for capturing, labeling and analysing.

Smart file types: A look at .tdf files.

• BTS smart system stores data in .tdf files.

• Files can be read using Smart analysissoftware and can contain kinematic, EMG,video, and various analog channels(pressure, sound, air flow, etc…)

• BTS also supply scripts for manipulation of.tdf files within the Matlab environment.

File Sharing (part I): Vicon to Smart.

• Source files: Vicon files of

clarinet player collected by Dr.

Wanderley’s lab.

• .c3d file containing channels

generated by plug-in gait was

unreadable.

• .c3d file containing only labeled

marker data was imported into

BTS Smart analyser.

File Sharing (part I): Vicon to Smart (con’t).

• C3d file was loaded and converted to .tdf

(see reconstruction)

• General Problems:

-rotation of reconstructed data occurs about

a fixed axis.

-Lab coordinate system rotated

-plug-in gait processed files are unreadable.

File Sharing (Part II): Smart to Vicon.

• Source files: Smart files of singer collected by Dr.

Cossette’s lab containing kinematics, video, EMG

and “sound”.

• Both plug-in gait and chest wall markers (128

markers!) used to collect and label data.

• Two files (full body and chest wall) were exported

separately using the BTS Smart c3d export tool.

File Sharing (Part II): Smart to Vicon (con’t)

• Full-body c3d:

-Trial was opened in Vicon workstation.

-Plug-in gait model applied

-kinematic variables generated.

• Chest-wall c3d

-Trial was opened in Vicon workstation

-89 markers were visible and labeled.

File Sharing (part II): Smart to Vicon (con’t).

• Problems

-use of two separate marker sets required

-System must be calibrated twice, once for

each marker set.

-video and emg not stored in .c3d.

Final Remarks

• Loss of channels: How best to export EMG

and video (mpg/avi)?

• Use of large number of markers requires

high camera count for accurate

reconstruction

• Combination of marker sets can be difficult.

• Different calibration volumes must be

calculated and stored.

Thank You!

With Support fromWith Support from

CFI , SSHRCCFI , SSHRC

•CIRMMT

•McGill Biomechanics lab

•Dr. Wanderley

•Erwin Schoonderwaldt