Download as ppt, pdf, or txt
Download as ppt, pdf, or txt
You are on page 1of 16

Similarity Analysis of Taekwondo Movement

Using Data Motion

Dharmayanti , Mohammad Iqbal, Adang Suhendra,


Achmad Benny Mutiara

Email : {dharmayanti , mohiqbal , adang,


amutiara}@staff.gunadarma.ac.id

1
INTRODUCTION
 Computer vision is a technology widely used to
overcome the limitations of human vision.

 Martial arts have a very fast dynamics of


movement, therefore it is important to be able to
digitize and analyze the movement.

 Motion capture data has great potential in sports


to measure athlete performance and measure the
specific needs required for movement.

2
INTRODUCTION Cont.

 Taekwondo is a Korean national martial art and one of


the martial arts competed in the Olympics as well as
various world competitions.

 The purpose of this study is developing an algorithm


that can identify the similarity of angular kinematic
motion in self-defense automatically with the kinematic
inverse method, fitting curve, and Pearson Correlation
Coefficients (PCC)

3
RESEARCH FRAMEWORK

4
Data Motion Acquisation

The acquisition of motion master and movement of


trainees using motion capture technology. Motion data
is stored in Bio Vision Hierarchy (BVH) format.

5
Pre-processing

 Preprocessing begins with reading the bvh file using a bvh


reading algorithm.

 The BVH files before being used for the analysis process are
filtered to eliminate interference during data capture.

 The data collection of [x, y, z] coordinate moves of each


joint and each frame, is stored in the form of a CSV file.

6
Determining Skeleton Articulation

7
Determine the x, y and z axis position
transformation with Inverse Kinematics
Methode
 Determining the change in the angle of
the motion position from the initial
position of the axis [x, y, z] to the axis
position [x, y, z] end-effector.

8
Determine velocity and acceleration of each
joint

The velocity of the position is the first derivative and the


acceleration of position using the second derivative.

Smoothing algorithm using fitting curve

To create algorithms to improve the graph of angular


results, speed and acceleration by doing fitting curve.

9
Algorithm Similarity Pearson Correlation
Coefficients (PCC)

 The PCC algorithm is used to measure the


linear correlation between the two datasets.

 The indicator of correlation is if the value is


in positive 1 or negative 1 but if the value 0
then declared not correlate

10
RESULT

Graphic x-axis position of hip joints Using PCC


Algorithm
11
RESULT

Graphic Velocity x-axis position of hip joints

12
RESULT

Graphic Acceleration x-axis position of hip joints

13
CONCLUSION

 This research can improve the processing of data


by producing algorithms for reading BVH and
analyzing Taekwondo movements automatically
and in real-time.

 The resulting algorithm has been able to produce


a measurement of the similarity of master motion
and trainees using PCC of 80% for the hip joint on
the x-axis, which means that the trainee moves
closer to the master movement on that axis.

14
CONCLUSION

 The results obtained between PCC and Cosine


Similarity are not much different, so it is necessary
to develop a method for analyzing motion.
 The PCC and Cosine Similarity cannot identify
movement mistakes made by trainees, only can
produce correlation values of similarity of motion
between master and trainee.
 Future research develops a method that can
analyze the similarity of motion and can also
identify errors in the motion.

15
THANK YOU

16

You might also like