post cta analyse for the ccp
TRANSCRIPT
Contour Analysis to Monitor Performance in Sports
Jeroen van der Eb
[email protected] 2015
Jeroen van der Eb. Snow 2015
Contour Analysis is a novel way for analyzing movements, developed for daily use in high performance sports. The Contour Analysis is part of an automaHc video capture and playback system, the CoachCockPit (CCP). Parameters like posiHon, velocity and average body angle are obtained and also less straighLorward parameters as step length or step frequency or 3D discus trajectory and speed can be invesHgated with Contour Analysis. The system is designed to operate fully automaHc with a minimum of user input. Performance parameters obtained from the Contour Analysis can be used for direct feedback and to monitor progress over longer periods of Hme. To illustrate this several examples in different sports are discussed. The CoachCockPit is developed with primarily the coach perspecHve in mind: it does not interfere with daily training. KEYWORDS: Markerless Movement Analysis, biomechanics, Performance Analysis, Feedback monitoring performance
Jeroen van der Eb. Snow 2015
Goal: Provide direct feedback to evaluate Performance for a daily training seYng for technical (individual) sports:
• Very easy to use • No interference with athlete • AutomaHc • Provide meaningful data • Quick feedback • Monitoring performance (session, week, year(s)) • Daily use • CompeHHon
In close collaboraHon with coaches and the support team dedicated modules (Post Analyzer) can be made for specific movements to provide quick feedback during training.
Jeroen van der Eb. Snow 2015
Contour Analysis (CtA)
Contour Analysis (CtA) obtains raw data form captured movements:
• Basics (raw) parameters: • (OpHcal) Center of Mass • Global body angle • Size • Length, width of body • …
• Other more complex parameters can be derived form the raw data: • Velocity of CoM • 3D trajectories of e.g. a ball or discus (using 2 (or more) synchronized camera’s)
• Stride length, frequency • Certain body angles • …
Jeroen van der Eb. Snow 2015
Principle of CCP and CtA
Contour Analysis (CtA) is part of the Coach Cockpit (CCP) which automaHcally captures and presents video clips of a movement. Typically 2-‐5 sec depending on the duraHon of the movement, but can be much longer for Cyclic movements. CtA is then automaHcally performed and relevant data is fed back to the athlete/coach. CtA provides raw data of the movement. De raw data has to be fed to a Post Processor to provide the coach/athlete with useful data. For a specific movement a dedicated Postprocessor has to made. Snow (Jeroen van der Eb) has the experience to develop a Post Processor for specific movement in close collaboraHon with coaches and the support team.
Jeroen van der Eb. Snow 2015
Specific CCP (generic)
Recorder Player CtA Post CtA Processor
CONTOUR ANALYSIS (CtA): Principles
Jeroen van der Eb. Snow 2015
1. Capture video 2. Contour Analysis 3. Biomechanical data
1
2
3
KinemaJcs of Vault Vault Height has a good correlaJon to the Judges Score (WC 2010)
Jeroen van der Eb. Snow 2015
11 12 13 14 15 16 170
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5Vertical velocity relative to score and type of vault. # 613
Jury Score
Vertical Velocity (m/s)
11 12 13 14 15 16 171.5
2
2.5
3
3.5
4
4.5
5Horizontal velocity relative to score and type of vault. # 613
Jury Score
Horizontal Velocity (m/s)
YurchenkoYurchenko 1/2HandspringTsukahara
11 12 13 14 15 16 170
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5Vertical velocity relative to score and type of vault. # 613
Jury Score
Vertical Velocity (m/s)
11 12 13 14 15 16 171.5
2
2.5
3
3.5
4
4.5
5Horizontal velocity relative to score and type of vault. # 613
Jury Score
Horizontal Velocity (m/s)
YurchenkoYurchenko 1/2HandspringTsukahara
11 12 13 14 15 16 17 Judges Score
VerHcal V
elocity
(m/s)
!
Num
ber o
f Vaults
Horizon
tal V
elocity
(m/s)
Max Vault Height versus Number of Twists 2 Twist 2.5 Twist
Center of Mass
Distance (m)
Height (m
)
2.3 2.4 2.5 2.6 2.7 2.8 Max Height of Vault (m)
Velocity at rake-‐off
Judges Score 11 12 13 14 15 16 17
Monitoring Performance using Vault Height
Jeroen van der Eb. Snow 2015
04/12/12 19/12/12 26/01/13 22/02/13 08/04/132.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
2.9
3
3.1
chronlogische volgorde gemaakte sprongen
hoog
te (m
)
YurchenkoTsukaharaOverslag
Vault H
eight (m)
Chronological order of Vaults
0 1 2 3 4 5 6 72.6
2.7
2.8
2.9
3.0
2.6
2.7
2.8
2.9
Sprung Nummer
Höh
e (m
)Va
ult H
eight (m)
Vault Trial number
One training session
Typical data form a training session. Vault height increases during warm-‐up (2-‐3 vaults) then 2 performance akempts. The gymnast stops when performance decreases (6th akempt)
Yurchenko Tsukahara Handspring
04/12/12 19/12/12 26/01/13 22/02/13 08/04/132.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
2.9
3
3.1
chronlogische volgorde gemaakte sprongen
hoog
te (m
)
YurchenkoTsukaharaOverslag
Progress
Hurdle: Performance analysis using Center of Mass 5 aSempts compared relaJve to the slowest aSempt
Jeroen van der Eb. Snow 2015
10 15 20 25 30 Distance (m)
Height of C
oM (m
)
Distance (m) Distance RelaH
ve to
Slowest T
rial (m)
In collaboraHon with R. van den Tillaar
Monitoring training load of elite trampolinists Using Angular and CoM data obtained from CtA
Example: 1 rouJne 27 jumps. The post processor automaJcally detects number of rotaJons and thus classifies the rouJne
Jeroen van der Eb. Snow 2015
Accuracy: The accuracy of the system is not easily given. It the depends on many factors, such as the specific movement and the group of athletes to analyze. The intenJon of the system is to provide quick and meaningful feedback to (elite) athletes without hampering the athlete. CtA is a trade of between user friendliness and number of outcome measures of a movement. For a specific movement applicaJon the accuracy has to be determined. Next: Implement in different sports. Combined with research to find valid performance parameters for specific sports Design proper presentaJon format of data for feedback for the specific need of the sport. For more informaJon:
Jeroen van der Eb [email protected]
Jeroen van der Eb. Snow 2015