quantification of arm joint motion for monitoring...
TRANSCRIPT
QUANTIFICATION OF ARM JOINT MOTION FOR
MONITORING BADMINTON PERFORMANCE
USING MEMS BASED WEARABLE SENSOR
ALVIN JACOB A/L VADANAYAGAM STEPHEN
UNIVERSITI TUN HUSSEIN ONN MALAYSIA
QUANTIFICATION OF ARM JOINT MOTION FOR MONITORING
BADMINTON PERFORMANCE USING MEMS BASED WEARABLE
SENSOR
ALVIN JACOB A/L VADANAYAGAM STEPHEN
A thesis submitted in partial
fulfilment of the requirement for the award of the
Degree of Masters of Electrical Engineering
Faculty of Electrical and Electronic Engineering
Universiti Tun Hussein Onn Malaysia
AUGUST 2017
iii
DEDICATION
“To almighty God....”
To my beloved parents,
Vadanayagam Stephen and Elizabeth Ghanam
For being the backbone of my life by supporting me from the very beginning
To my supervisor,
Dr Wan Nurshazwani Wan Zakaria
For their consistent encouragement, guidance and support throughout the research
To my siblings,
Adrian Sarron and Aaron Matthew
For their trust and motivation during my studies
To my friends
For their support and encouragement through all these years
Who understand and guided me through the journey of my study.
And for all who believe in me and keep whispering
“You can do it”
iv
ACKNOWLEDGEMENT
First, above of all, I would like to take this opportunity to express my gratefulness to
Almighty God for the guidance and knowledge that he has granted throughout life’s
journey. With all His blessing, I successfully managed to complete my studies.
My greatest appreciation goes to my beloved family especially my parent
who had given me support in term of moral and financial, throughout my study in
University Tun Hussein Onn Malaysia. In addition, my brothers who had given me
the courage and strength that are necessary for me to carry on with my studies.
I would like to convey my gratitude to my supervisor, Dr Wan Nurshazwani
Wan Zakaria who had guided and supported me throughout the completion of my
studies. I am very grateful for the financial assistance provided throughout the
journey of this research. Without her critics and valuable suggestions, this research
project and thesis writing would have been difficult to be completed in time.
Lastly, my sincere appreciation also extends to all either contributed directly
or indirectly especially to all my colleagues for their cooperation and unflagging
support to complete this research successfully.
Alvin Jacob A/L Vadanayagam Stephen
Batu Pahat, Johor
v
ABSTRACT
Recent advancement in wearable sensor technology has made a significant impact in
the field of sports performance analysis, as it dramatically changes the conventional
ways of training athletes. Sensor based systems help coaches to monitor, interpret
and analyse athlete’s performance, where technology is used to train professionals.
This study aims to develop a Wearable Sensor System (WSS) to monitor and
quantify a badminton players arm motion, especially the wrist and elbow joint when
performing serve and drive strokes. Current monitoring systems are unable to
classify player’s competency level, due to insufficient technology to quantify joint
variables and subsequent analysis on related badminton movement. However,
previous studies suggest the vision-based systems, either using high-speed cameras
or pre-recorded video analysis; but both systems are costly and are inconsistent in
monitoring a badminton player due to environmental conditions (excessive lighting).
Therefore, two monitoring modules are developed for the WSS. First, Hand Wrist
Monitoring Module (HWMM) that uses flex sensors to measure player’s finger
flexion and Inertia Measurement Units (IMU) to measure wrist rotation angle.
Second, Elbow Monitoring Module (EMM) that used one IMU sensor to measure
elbow joint rotation angle. Additionally, Real Time Operating System (RTOS) is
implemented to manage simultaneous sensor data acquisition for synchronisation
between monitoring modules. As a result, the WSS provides reliability and accuracy
up to 95% for static joint motion measurements. Encouragingly, 6 out of 19 players
were found to have improvements in their serve technique, whereas 12 players were
able to perform a drive stroke with more than 80% similarities with the coach.
Therefore, these findings emphasise the importance of a monitoring tool to help
improve a badminton player’s performance, where it contributes by enhancing the
coaching process by providing statistical and quantified movement data.
vi
ABSTRAK
Kemajuan terkini dalam teknologi penderia boleh pakai telah memberi impak yang
besar dalam bidang analisis prestasi sukan, kerana ia secara dramatik mengubah cara
konvensional atlet dilatih. Sistem berasaskan penderia membantu jurulatih untuk
memantau, mentafsir dan menganalisis prestasi pemain. Oleh it, kajian ini
membangunkan Wearable Sensor System (WSS) untuk memantau dan mengukur
pergerakan lengan pemain badminton, terutamanya pergelangan tangan dan siku.
Sistem pemantauan semasa tidak dapat mengklasifikasikan tahap kompetensi
pemain, kerana tiada teknologi yang mencukupi untuk mengkuantifikasi
pembolehubah sendi dan menganalisis pergerakan badminton yang berkaitan. Walau
bagaimanapun, kajian lepas mencadangkan penggunaan sistem penglihatan, tetapi
sistem ini adalah mahal dan tidak konsisten dalam memantau pergerakan pemain
badminton disebabkan oleh keadaan persekitaran (pencahayaan yang berlebihan).
Untuk itu, dua modul pemantauan telah dibangunkan untuk WSS. Pertama, Hand
Wrist Monitoring Module (HWMM) yang menggunakan sensor flex untuk mengukur
sudut fleksi jari pemain dan Inertia Measurement Units (IMU) untuk mengukur
sudut gerakan pergelangan tangan. Kedua, Elbow Monitoring Module (EMM) yang
menggunakan satu sensor IMU untuk mengukur sudut gerakan siku. Di samping itu,
Real Time Operating System (RTOS) telah diimplementasikan bagi menguruskan
pemerolehan data sensor. Hasilnya, WSS memberikan kebolehpercayaan dan
ketepatan sehingga 95% untuk pengukuran pergerakan statik. Didapati bahawa, 6
dari 19 pemain menunjukkan peningkatan dalam teknik serve mereka, manakala 12
pemain mampu melakukan pukulan drive dengan persamaan sebanyak 80% dengan
jurulatih. Oleh itu, penemuan ini menekankan pentingnya alat pemantauan prestasi
bagi membantu meningkatkan prestasi pemain badminton, di mana ia boleh
menyumbang kepada penambahbaikan proses melatih pemain dengan menyediakan
data pergerakan statistik dan kuantitatif.
vii
TABLE OF CONTENTS
TITLE i
DECLARATION ii
DEDICATION iii
ACKNOWLEDGEMENT iv
ABSTRACT v
ABSTRAK vi
TABLE OF CONTENTS vii
LIST OF FIGURES x
LIST OF TABLES xiv
LIST OF SYMBOLS AND ABBREVIATIONS xv
LIST OF APPENDICES xvi
LIST OF PUBLICATIONS xvii
CHAPTER 1 INTRODUCTION 1
1.1 Introduction 1
1.2 Problem statement 4
1.3 Research Contribution 5
1.4 Aim 5
1.5 Objectives 5
1.6 Research Hypothesis 6
1.7 Scope and limitation 6
1.8 Thesis Outline 6
CHAPTER 2 LITERATURE REVIEW 8
2.1 Sports Performance Measurement and Evaluation 8
2.2 Current Technology on Wearable Sports Measurement
Methods 10
2.2.1 Vision-Based Motion Capture Devices 11
viii
2.2.2 Textile-Based Devices 17
2.2.3 Flexion Based Monitoring System 23
2.2.4 MEMS Based Motion Capture Devices 26
2.3 Current Technology on Wearable’s Communication 36
2.4 Wearable Monitoring Technology 38
2.5 Summary 39
CHAPTER 3 FEASIBILITY STUDY 41
3.1 Kinematic Analysis of the Human Arm 42
3.1.1 Kinematic of Hand 43
3.1.2 Kinematic of Wrist 44
3.1.3 Kinematic of Arm 45
3.2 Badminton Grip 46
3.2.1 Importance of a Proper Grip 47
3.2.2 Fundamental Badminton Grips 48
3.3 Movement in Badminton 51
3.4 Fundamental Badminton Stroke 52
3.4.1 Serve Stroke 52
3.4.2 Drive Strokes 55
3.5 Summary 56
CHAPTER 4 DESIGN AND DEVELOPMENT OF WEARABLE
SENSOR SYSTEM 58
4.1 Wearable System Design Criteria 58
4.2 Overall Monitoring System Design 59
4.3 Mobile Measurement Device Hardware Configuration 61
4.3.1 Hand Wrist Monitoring Module (HWMM) 61
4.3.2 Elbow Monitoring Module 73
4.4 Wearable Sensor Data Acquisition 76
4.4.1 Flex Sensor Data Acquisition 77
4.4.2 RTOS Synchronization 79
4.4.3 IMU Data Acquisition 81
4.5 Preliminary Tests 87
4.6 Design of Experiment 88
4.7 Summary 89
ix
CHAPTER 5 RESULTS AND ANALYSIS 90
5.1 Sensor Performance and Reliability 90
5.1.1 Flex Sensor 90
5.1.2 Inertial Measurement Units 96
5.1.3 Communication Tests 101
5.2 Human Motion Measurement 105
5.2.1 Wrist IMU 105
5.2.2 Elbow IMU 110
5.2.3 Grip Type 112
5.3 Experimental Programme 118
5.3.1 Design of Experiment 118
5.3.2 Serve Stroke 119
5.3.3 Drive Stroke 127
5.4 Survey Results 133
5.4.1 Personal Details 133
5.4.2 Developed Hardware 136
5.4.3 Overall Experience 140
5.5 Summary 142
CHAPTER 6 CONCLUSION AND RECOMMENDATION 144
6.1 Overview 144
6.2 Achievements 144
6.3 Future Works 146
REFERENCES 148
Vitae 148
x
LIST OF FIGURES
1.1 The Game of Poona 1
2.1 Process flow for performance measure 10
2.2 An actor is wearing a suit with passive reflective markers 12
2.3 A student wearing a suit with active markers illuminated with
red LED in Virtual Reality Lab of University of Texas 12
2.4 Elbow-shoulder alignment of a spin bowl 13
2.5 (a) Attached markers and (b) MOCAP reanimation data 14
2.6 Configuration of the stereo camera and IMU sensor 14
2.7 The placement of markers on test subject 15
2.8 Smash motion for ground shuttle 16
2.9 Smash motion for in-air shuttle 16
2.10 Knitted stretch sensor (a) and the (b)sensor jacket final design 18
2.11 Novel fabric-based sensor design 19
2.12 Textile with (a)embedded electrodes and (b)snap connector 20
2.13 The developed fabric sensor for (a) elbow 21
2.14 (a) The waistband with (b) integrated pH sensor pocket and (c)
experimental set-up 22
2.15 Chest band to measure breathing rate 22
2.16 Full finger flexion (0), half extension (1) and full extension (2) 23
2.17 Developed data glove 24
2.18 Glove integrated with flex sensor for (a)hand rehabilitation and
(b) sensorized ball 25
2.19 Block diagram for motion tracking device 27
2.20 Kinematic chain of lower limbs and IMU sensor placement 28
2.21 Overall system configuration for lower limb tracking 29
2.22 Jogging velocity 30
2.23 Sketch map of the walking motion 31
xi
2.24 Major component of the developed wireless MEMS 31
2.25 Placement of IMU sensor 32
2.26 Mobile measure device sensor placement 33
2.27 Sensor position along human arm 34
2.28 Sensor placement for smash stroke tracking 35
3.1 Human hand skeleton model (a) and kinematic configuration of
the human hand (b) 43
3.2 The wrist types of motion 44
3.3 Human arm joint illustration (a) and range of motion for each
joint (b) 45
3.4 Forehand and backhand grip illustrations 48
3.5 Panhandle grip, front view (a) and back view (b) 49
3.6 Thumb grip, front view (a) and back view (b) 50
3.7 Bevel grip, front view (a) and back view (b) 51
3.8 Stroke cycle (a) and range category (b) 51
3.9 Serve Stroke 52
3.10 Low serve trajectory 54
3.11 High serve trajectory 54
3.12 Drive shot trajectory 55
3.13 Forehand drive 55
3.14 Backhand drive 56
3.15 Sensor orientation and placement on the human arm 57
4.1 The final design of monitoring system for badminton player 60
4.2 Hand monitoring module (HMM) 62
4.3 Wrist monitoring module (WMM) 63
4.4 Bend sensor by Spectral Symbols 64
4.5 First HMM prototype 67
4.6 Flow diagram of the master and slave pairing process 69
4.7 First HWMM prototype 71
4.8 HWMM circuit 71
4.9 Final design of the HWMM 72
4.10 Elbow monitoring module (EMM) 73
4.11 EMM circuit 74
4.12 The EMM main controller board and pouch 75
xii
4.13 The final design of EMM 75
4.14 Graphical user interface for the WSS 77
4.15 Flex sensor data acquisition flow chart and pseudo code 78
4.16 General RTOS task state 80
4.17 Priority based task pipelining for HWMM 80
4.18 IMU sensor data acquisition flow chart 81
4.19 Library function (1) for IMU Sensor 82
4.20 Library function (2) for IMU Sensor 83
4.21 Library function (3) for IMU Sensor 84
4.22 Library function (4) for IMU Sensor 85
4.23 IMU sensor data acquisition flow chart (continued) 86
5.1 Flex sensor resistance mean value representation 92
5.2 Comparison between first and second test resistance value 94
5.3 RC filter circuit for the flex sensor 95
5.4 Flex sensor test procedure 95
5.5 Output voltage against measured angle 96
5.6 IMU Sensor axis aligning 97
5.7 IMU testing circuit 97
5.8 IMU sensor calibration procedure 98
5.9 IMU sensor testing procedure 100
5.10 Bluetooth connection timing diagram 102
5.11 Measured time against measured angle 102
5.12 RTOS cooperative context task switching 103
5.13 RTOS measurement for index and middle finger 104
5.14 RTOS measurements for HWMM 105
5.15 Orientations of the human hand (a), (b), (c) and (d) 106
5.16 Elbow (a) flexion and (b) extension 110
5.17 Forehand grip illustration 113
5.18 Forehand finger flexion angle 113
5.19 Backhand grip illustration 114
5.20 Backhand finger flexion angle 114
5.21 Panhandle grip illustration 114
5.22 Thumb grip illustration 115
5.23 Bevel grip illustration 115
xiii
5.24 Grip finger flexion angles 116
5.25 Loose grip 117
5.26 Tight grip 117
5.27 Coach serve stroke 120
5.28 Coach wrist rotation angle 120
5.29 Player 1 serve stroke 121
5.30 Player 1 wrist rotation angle 121
5.31 Player 3 serve stroke 123
5.32 Player 3 wrist rotation angle 123
5.33 Player 6 serve stroke 124
5.34 Player 6 wrist rotation angle 124
5.35 Player 9 serve stroke 125
5.36 Player 9 wrist rotation angle 125
5.37 Low serve test procedure 126
5.38 Coach drive stroke 129
5.39 Coach elbow rotation angle 129
5.40 Player 8 drive stroke 130
5.41 Player 8 elbow rotation angle 130
5.42 Player 9 drive stroke 131
5.43 Player 9 elbow rotation angle 131
5.46 Drive shot test procedure 132
5.47 Players faculty 134
5.48 Players age and gender 135
5.49 Players experience level 135
5.50 Players skill level and dominant hand 136
5.51 Monitoring module user-friendliness 137
5.52 Monitoring module design 137
5.53 Glove sensors visibility 138
5.54 Monitoring module usage experience 138
5.55 Monitoring module future usage 139
5.56 Sports monitoring device future usage 140
5.57 Monitoring device investment 140
5.58 Continues development 141
5.59 Overall experience 142
xiv
LIST OF TABLES
2.1 Types of MOCAP technique 12
2.2 Wireless communications standards 37
2.3 MOCAP Technology Overview 40
2.4 Badminton MOCAP Review 40
3.1 Six stages of performance evaluation 42
3.2 Range of motion (degree) of the wrist 44
3.3 Range of motion (degree) of the elbow 46
4.1 MPU6050 IMU sensor properties 65
4.2 Comparison between the HMM and HWMM prototype 70
5.1 Resistivity to angle measurement 91
5.2 Result of resistance to voltage conversion 93
5.3 IMU sensor raw data 99
5.4 IMU calibration results 101
5.5 Results for hand orientations 107
5.6 Results of wrist flexion and extension 108
5.7 Results of wrist radial and ulnar deviation 109
5.8 Results of elbow flexion 111
5.9 Result of elbow extension 112
5.10 Maximum and minimum grip values 117
5.11 Serve stroke accuracy test results 127
5.13 Drive stroke accuracy test results 132
xv
LIST OF SYMBOLS AND ABBREVIATIONS
IBF - International Badminton Federation
UTHM - Universiti Tun Hussein Onn Malaysia
MEMS - Microelectromechanical Systems
GUI - Graphical User Interface
IMU - Inertia Measurement unit
MM - Monitoring Module
MIMU - Miniature Inertial Measurement Unit
MMD - Mobile Measuring Device
LED - Light Emitter Diode
MOCAP - Motion Capture
I2C - Inter-Integrated Circuit
ADC - Analogue Digital Converter
YPR - Yaw, Pitch and Roll
BSN - Body Sensor Network
WSS - Wearable Sensor System
WSMS - Wireless Sports Monitoring System
WSN - Wireless Sensor Network
HMM - Hand Monitoring Module
WMM - Wrist Monitoring Module
EMM - Elbow Monitoring Module
HWMM - Hand Wrist Monitoring Module
xvi
LIST OF APPENDICES
A Flex Sensor 158
B Flex sensor test data 159
C Bluetooth communication test data 160
D HWMM design and development 161
E MPU6050 IMU Sensor 163
F Flex Sensor Program Code 164
G MPU6050 Library 165
H HC-05 Bluetooth Module 169
I RTOS Code 170
J Survey Form 172
K Student Name List 178
L Players Group Analysis Data 179
xvii
LIST OF PUBLICATIONS
Jacob, A., Zakaria, W. N. W., & Tomari, M. R. B. M. (2015). Implementation of
Bluetooth Communication in Developing a Mobile Measuring Device to
Measure Human Finger Movement. ARPN Journal of Engineering and Applied
Sciences, 10(19), 8520–8524.
Jacob, A., Zakaria, W. N. W., & Tomari, M. R. B. M. (2016). Evaluation of I2C
Communication Protocol in Development of Modular Controller Boards. ARPN
Journal of Engineering and Applied Sciences, 11(8), 4991–4996.
Jacob, A., Zakaria, W. N. W., & Tomari, M. R. B. M. (2016). Implementation of
IMU sensor for elbow movement measurement of Badminton players. In IEEE
(Ed.), 2016 2nd IEEE International Symposium on Robotics and Manufacturing
Automation (ROMA) (pp. 1–6). Ipoh: IEEE.
Jacob, A., Zakaria, W. N. W., & Tomari, M. R. B. M. (2016). Quantitative Analysis
of Hand Movement in Badminton. Lecture Notes in Electrical Engineering (Vol.
362, pp. 439–448). Phuket: Springer International Publishing.
CHAPTER 1
INTRODUCTION
1.1 Introduction
Badminton is a modern version of an ancient game which was played around 2000
years back in the ancient Greece, China, and India where it was called the battledore
(bat or paddle) and shuttlecock (Boga, 2008). Back in the 1600s, the Battledore and
Shuttlecock was an upper-class pastime in England and many European countries
where these two games were played by two people simply hitting a shuttlecock
backwards and forwards with a single bat as many times as they could without
allowing it to hit the ground (Guillain, 2004). Later in the 1800s, the game was called
‘Poona’ or ‘Poonah’, which was played throughout India, this version of the match
had a net in between and players hit the shuttlecock across the net (Guillain, 2004).
Figure 1.1 illustrates how to play a game of ‘Poona’.
Figure 1.1: The Game of Poona (Guillain, 2004)
2
In the mid-1800s the British officers stationed in India took this game back to
England. After being adopted and played, badminton became popular in England and
the first badminton championship match held in 1898. Then in 1934, the
International Badminton Federation was formed, with the first members including
England, Wales, Ireland, Scotland, Denmark, Holland, Canada, New Zealand and
France, with India joining as an affiliate in 1936 (Federation, 2006). In 1966,
badminton was added to the Commonwealth Games and later in 1992 introduced to
the Olympic Games.
Badminton is a fast and dynamic sport that can be classified as one of the
fastest racket sports, with the fastest recorded badminton hit in a competition at 408
km/h achieved by Lee Chong Wei from Malaysia in 2015 (Guinness World Records,
2015). Movements performed by badminton players are a combination of different
type of motions that consists of hard smashes, short drop and long clears where all
these movements force the player to act and react in an extremely rapid manner.
About 20% of the attacks performed during a game are smashes, and jump smashes
(Tong & Hong, 2000). Badminton is played by anyone regardless of the age or
experience; professional players spend hours training to improve their playing skills.
Proper playing technique is crucial, and it is the deciding factor between
winning and losing a badminton game. Experienced badminton players have
different stroke techniques and faster stroke rate compared to amateur players or
beginners which are results of continues training. However, it can be a tedious and
long process for beginners just to sit back and watch the professional players play to
discover the exact difference in the motion performed. This is because of the
difference in movement speed and strokes complexity demonstrated by the
professional players. This data is important and can be used to determine player’s
competence level. Coaches can improve their teaching and training methods if the
competence level of the player is known beforehand.
Scientific research on tactics, strategy, or playing patterns of international
level badminton players are, however very limited and restricted by confidentiality
issues (Tong et al., 2000). On the contrary, researchers have been conducting
research focusing on measuring the speed of the badminton smash stroke, which is
pigeonholed as the winning shot. This statement is supported by the research carried
out by Salim et al. (2010) to identify the difference of forehand overhead smash
performed by male and female players. The researchers concluded that the upper arm
3
rotation contributes the most to achieving a rapid smash and that male subject
reaches higher racket velocity faster than female subjects do. Similarly, another
research was performed to investigating the upper limb movement during a
badminton smash conducted by Shan et al. (2015), where the author concluded that
the wrist movement contributes the most when performing a badminton smash.
Since badminton, strokes are fast paced; it is certain that the human eye alone
is not capable of interpreting the rapid movement of a badminton player. Generally,
high dynamics movements are usually analysed using high-speed optometric systems
such as high-speed video as demonstrated by Salim et al. (2010) in their research.
They conducted arm motion analysis for smash stroke on male and female players,
by using four Oqus high-speed cameras by Qualisys (Qualisys, 2004) to record the
markers placed on the players. However, due to the technical limitations (e.g. high
amount of light) that exist when using this method, these measurements are often
performed in a laboratory setting. Hence, this does not comply well with the real
competition or training conditions (Jaitner & Gawin, 2010). Furthermore, Optometric
systems are expensive and require high-speed equipment’s to operate.
As a result, researchers began experimenting with the development of
electromechanical based speed and orientation sensors to overcome previous
limitations (Fong et al., 2004; Senanayake, 2004; Zatsiorsky, 1998). The miniature
sensors allow data collection at high sample rates with wider measuring range when
connected to a microcontroller. Furthermore, the light weight and small size of the
sensors does not restrict the performance of the player (Jaitner & Gawin, 2007).
Therefore, based on previous study comparisons, the best method suggested to
successfully capture a badminton players motion data while still offering flexibility
and keeping the cost low is by using wearable sensors.
This concept has been implemented and is being used in other sports type, for
example in golf by King et al. (2008) where MEMS-based accelerometer has been
used to measure the speed of a golf club swing. Wearable sensors are becoming
popular in many fields to measure human movement and dynamics, where it can be
used to monitor the body’s physiological response and the kinematic aspects of
performance. Continues performance monitoring is the key for players to improve in
the court, to monitor this in a natural way there is a need for integrated sensors that
are straightforward to use and comfortable to wear (Coyle et al., 2009). The
measurement and evaluation process is necessary to make players feel motivated to
4
perform better, and it helps coaches to analyse and improve the performance of their
players (Henao & Fruett, 2009; Morrow et al., 2005).
1.2 Problem statement
The ability to acquire and develop skills is very crucial for badminton, where the
difference between winning and losing is often determined by the level of skill and
techniques utilised by the players. Although, badminton is one of the most played
racket sports in Malaysia; however, with the lack of proper and inadequate training
the younger generation seems to be having problems in acquiring professional-level
skills. In an interview session with a Badminton coach, the coach expressed that
there is no specific tool to measure the skill and competency level of the players
except for visual classification by the coach or players them self (Sazali, personal
communication, July, 2015; Shan et al., 2015).
The coach stated that by knowing the competency level of each player, it
would be easier to categorise them into groups of beginners, intermediate and
advance level players. It is essential, as specialised training programs could be
designed and provided to improve players skill sets (Shan et al., 2015; Sen et al.
2015). However, a scientific study carried out to perform this classification are very
less as most study focuses on badminton stroke analysis of professional players. In
fact, numerous studies have been conducted to measure and analyse the properties of
a smash stroke, where the analysis focusses more on the shoulder and elbow joint
(Jaitner et al., 2010; Kiang et al., 2009). Due to this, it is not practical to use the
system to analyse the wrist joint. Moreover, some researchers have conducted studies
to identify the type badminton stroke used by players; however, they all focus on the
stroke analysis rather than on the execution.
One of the major problem for beginner level players is the incorrect way of
gripping the badminton racket (Sazali, personal communication, July, 2015).
Therefore, a method to classify the recommended grip method for each badminton
stroke is required. However, to the best of author’s knowledge, no research has been
carried out so far to perform this classification. As a result, the grip type
identification is a novelty to the proposed monitoring system. Furthermore, current
systems are mostly based on vision tracking systems, which requires the use of high-
5
speed camera and high processing power computers. However, they are costly,
require longer set up time and is limited to laboratory conditions. Thus, it is
necessary to conduct a research to overcome aforementioned problems.
1.3 Research Contribution
Most of the current research related to badminton performance measurement is
focused mainly on the smash stroke and players playing pattern analysis. However,
very little attention has been given to the methods required to classify 1) players
according to competency level and 2) different badminton strokes. Therefore, this
research provides insight on the method used to quantify player’s movement mainly
for serve and drive badminton strokes. Moreover, this is the first attempt made to
identify and classify the different grip types used by badminton player. Finally, the
proposed Wearable Sensor System (WSS) focuses on establishing a comprehensive
individual database of the player’s movement data. This research provides coaches
with a monitoring tool for performance analysis to help and motivate players to
perform better.
1.4 Aim
The primary aim of this research is to develop a Wearable Sensor System to assist in
monitoring badminton player’s performance.
1.5 Objectives
The objectives of this study are:-
1) To assess the fundamental joint movements; 1) hand, 2) wrist and 3)
elbow of a badminton player.
2) To develop a Wearable Sensor System (WSS) to quantify different grip
types and badminton strokes based on identified joint movement.
3) To analyse and evaluate the performance of the developed WSS via a
series of test and experimental program.
6
1.6 Research Hypothesis
The developed WSS could have a significant impact on quantifying a badminton
player’s movement to increase their competency level.
1.7 Scope and limitation
The scope and limitation of this project are as follows:
1) The targeted area for this research is badminton singles.
2) This research focused on two badminton strokes, which is the service, and
drive stroke.
3) The measurement is limited to three parameters, which are; finger flexion
angle, wrist and elbow rotation angle.
4) The experimental study involved 19 badminton players age between 22 to
26 years old.
5) The developed WSS is validated using the following equipment’s, 1)
goniometer, 2) Angle protractor and 3) Digital Multimeter.
1.8 Thesis Outline
This thesis is organised as follows:-
Chapter 1 gives an introduction to badminton sport and an overview of sports
monitoring systems that are being used. Also, the research background, problem
statement, objective and scopes are presented.
Chapter 2 discusses the related work that has been done in sports monitoring area
and presents the current technology being used to measure sports performance.
Moreover, examples of performance measurement and communication technology
used are also discussed.
7
Chapter 3 elaborates in-depth about the kinematic analysis of the human arm related
to badminton stroke and also investigates the major strokes used in badminton.
Chapter 4 discussed the development process of the WSS and the overall
monitoring system. Furthermore, the development of the individual sensor modules
is presented.
Chapter 5 elaborates on the experiments designed to test the developed WSS
system, where the results obtained from the system is analysed by making a
performance comparison between the players and coach.
Chapter 6 summarises the presented work and outlines the conclusions of the
results. Moreover, the research contribution and future work related to this research
are discussed.
CHAPTER 2
LITERATURE REVIEW
Recent advancements in wearable sensor technology have made a significant impact
in the field of sports technology, where it has influenced the development of many
specific sports training equipment’s used to enhance an athlete’s performance.
Therefore, the conventional ways of training an athlete are dramatically changed
with the help of technology. A review of the recent technologies currently used in
sports has been performed to acquire information related to the research subject.
Importantly, this information and methodologies are used as guidelines for selecting
suitable and appropriate features for the proposed monitoring system development
and the data acquisition. The monitoring system, with which athletes can be
monitored and analysed; would help athletes in developing skill and performance in
their respective area. Briefly, this research is trying to develop more badminton
professionals with the aid of technology.
2.1 Sports Performance Measurement and Evaluation
Performance measurement in sports is a process of collecting and analysing
information concerning the performance of an individual or a group. It involves
studying and analysing processes to determine whether output received are in line
with what was intended or should have been achieved. This research aims to quantify
the movement data of a badminton coach using the developed monitoring system to
help players to learn and improve their skills based on the data, where players and
coaches can compare data to understand if they are in line. Analysing and evaluating
the measured data, the difference between a normal player and a professional player
9
should be noticeable. One major difference between a normal and a professional
player can be observed by the level of dedication and effort expressed by the players
towards training. Evaluation is a process of making decisions based on statements of
quality, merit or value about what is being assessed (Morrow, Jackson, & Disch,
2011). Once an athlete’s performance is evaluated based on how well a task is
executed, training programs designed to enhance the ability of the athlete largely
dependents upon satisfying the performance aims associated with it (MacKenzie,
1997). Though sport performance measurements are necessary, questions remain on
why should it be performed or evaluated. According to Morrow et al. (2011), there
are six general purposes for performing these measurements:
1) Placement - Grouping players according to their ability based on testing and
evaluations done by a professional trainer speeds up the training process, as
they have a similar starting point and can improve consistently.
2) Diagnosis - Test results provide plenty information about the physiology
aspect of a player. Evaluating these results enables the coach to determine the
weakness or deficiencies of a player to help them perform better.
3) Prediction - One of the main goals for measuring human performance is to
be able to use present or past data to predict the future outcomes.
4) Motivation - Test results can be used as encouragement to achieve more;
where players need the stimulation and feedback from their evaluation to be
able to achievements and perform better.
5) Achievement - Improving human performance is important in a training
program, where players strive to achieve a predetermined goal.
6) Program Evaluation - Measurement data taken using players from different
competence level of the same training program can be compared with data
from another training program to validate the effectiveness of the training
program.
10
Training programs are designed to train athletes and they have predefined
goals for the athletes to achieve, this motivates them individually to perform better.
As mentioned beforehand, this research aims to develop a WSS that is capable of
monitoring the moves of a badminton player. The collected data helps coaches to
place players into groups with same skills level, which would speed up training time
and help players to develop self-confidence. Coaches play an important role, as all
the testing and measurements done; are just ways of collect information. Moreover,
the data is used as the baseline for performance evaluation and decision making on
the types of training to be provided. Hence, knowing the reason why performance
measurements are done, next is to investigate how these measurements are
performed and monitored.
2.2 Current Technology on Wearable Sports Measurement Methods
The motivations to measure an athlete’s performance may depend on the six reasons
discussed beforehand, but a professional coach or a trainer determines the parameters
where these performance measurements are conducted. According to Bartlett (2007),
there are three essential parts in taking a measurement as shown in Figure 2.1. First,
there should be a method or a device used for measuring the parameter needed.
Second, a system that interprets all the data obtained. This system may in the form of
a human being or a computer system that interprets the raw data obtained from the
measuring devices to provide feedback. Third, a method to provide feedback that
could be in the form of visual or practical training. Baca & Kornfeind (2006) quotes
that “elite sports training programs use rapid feedback systems to acquire and present
relevant performance data shortly after motion execution using embedded sensors
and devices.” Thus, the present study investigates how basic human movement is
recorded and what type of measurement method researchers use to measure athlete
performance from various forms of sport.
Figure 2.1: Process flow for performance measure
Device Interpreter Feedback
11
2.2.1 Vision-Based Motion Capture Devices
In the 1970s and 1980s, the photogrammetric analysis was used as motion capture
tool in research (Generation, 1995). Later around the 20th century, usage of video
capture devices was introduced; where it could capture the continuous motion of an
object (Fischer, 2013). For example, research conducted by Tong & Hong (2000)
uses taped video of total ten recorded match from the 1996 Hong Kong Badminton
Open Tournament to analysis the percentage distribution of different shot types
utilised by players in a badminton game. Interestingly, the use of computers to
analyse human motion is gaining popularity in the field of motion capture. Moreover,
the most important part of motion capture is the task of registering the motion
performed, a process known as human motion capture (Moeslund & Granum, 2001).
Motion capture covers many aspects, but it is primarily used in capturing
large scale body movements, which include movements of the head, arms, torso, and
legs (Moeslund et al., 2001). Vision-based Motion Capture is a large topic and has
been popularly used through the world in many fields (Moeslund, Hilton, & Krüger,
2006). However, in the recent decade; it has been employed in the animation and
movie industries especially to produce Computer-generated imagery (CGI). Motion
Capture (MOCAP) is the process of recording the movements of a subject or an
object, whereas it is often called performance capture when a living subject is
involved (Noiumkar & Tirakoat, 2013). Many research has been conducted by
integrating this technology into the development of better sports equipment’s, which
proved to be beneficial to the sports industry (Mirabella et al., 2011; Nam, Kang, &
Suh, 2014; Shingade & Ghotkar, 2014).
In sports, vision-based MOCAP technique is usually used to create animation
model of a human performing certain movement, where the data obtained is
reanimated into a 3D model for data analysis (Noiumkar et al., 2013). Table 2.1
shows two basic MOCAP techniques that are used in sports. Vision-based MOCAP
system uses data captured from image sensors of two or more camera to triangulate
the 3D position of a subject, where reflective markers provide tracking algorithm
with points to build a skeleton model of the subject (Moeslund et al., 2001).
Advancements in marker technology have enabled MOCAP systems to generate
accurate data by tracking surface features (Moeslund et al., 2006).
12
Table 2.1: Types of MOCAP technique
Passive Markers
The subject wears a suit fitted with
retro-reflective coated markers as
shown in Figure 2.2. Cameras emit
light, which is reflected by these
markers to detect the position of the
marker. The intensity of the light
produces is controlled so that only
the reflective markers are sampled,
eliminating skin and fabric
(Generation, 1995).
Figure 2.2: An actor is wearing a suit with
passive reflective markers
Active Markers
The subject wears a suit fitted with
LED-illuminated markers; these
markers are tracked using high-
speed cameras for the position of
the wearer as in Figure 2.3. The
illuminated marker results in lower
marker jitter and an increase in high
measurement resolution that allows
the system to track the marker
position precisely. Some systems
have multi coloured LED markers
to distinguish different parts on the
human body
Figure 2.3: A student wearing a suit with
active markers illuminated with red LED in
Virtual Reality Lab of University of Texas
(Johnson, 2011)
The two different MOCAP method shown in Table 2.1 are popularly being
used in many research to record and map human motion into 3D models, where the
model can then be analysed to extract further information like movement angles,
speed and body physiology. For example, Montazerian et al. (2008) used passive
13
markers to capture the 3D animation of a full ball delivery action by the Imperial
College cricket team bowlers. The goal was to obtain the 3D kinematics of the
bowler's elbow angle during the ball swing. Therefore, Euler angles for the arm
rotation was calculated in the sequence of zx’y” where z is flexion-extension, x is
abduction-adduction and y is the internal-external rotational. Figure 2.4 shows the
graphically rendered bowler's upper torso skeleton model, where the motion is
represented by obtaining the position of markers.
Figure 2.4: Elbow-shoulder alignment of a spin bowl (Montazerian et al., 2008)
Noiumkar & Tirakoat (2013) conducted a research using the same passive
marker MOCAP method to record the motion of hitting a golf ball. The researchers
integrated optical MOCAP technology to capture the swing stroke motion to
represent and remodel it into a 3D model. Therefore, Noiumka et al. (2013) used
three different MOCAP marker set configurations, which is optimised 24-point,
animation 28-point, and normal 42-point. The 3D character model animation is
rendered from the data obtained from the three marker sets, where the results indicate
the optimised 24-point and animation 28-points were equally good and is suitable to
be used to model the golf player. Whereas, the 42-point rendering required more time
and resources and showed delays for real-time applications. The transaction from the
motion data skeleton model to the fully rendered 3D character modelling of the
player is shown in Figure 2.5. The researcher concluded that the usage of optical
MOCAP is suitable to capture the movements in sports given that the suitable marker
optimisation method is used to track the movement of players.
14
Figure 2.5: (a) Attached markers and (b) MOCAP reanimation data (Noiumkar et
al., 2013)
In another research, Nam et al. (2014) utilised the same concept by using
optical MOCAP to track a golf club swing motion. Although the MOCAP method is
the same as the previous research, the marker technology used is different. For
instance, Nam et al. (2014) used active markers where an infrared LED illuminates
each marker. Eight infrared LED’s are mounted on a “T” shaped panel, which is
attached to the lower part of the golf club as shown in Figure 2.6.
Figure 2.6: Configuration of the stereo camera and IMU sensor (Nam et al., 2014)
Additionally, the marker set is tracked using two stereo cameras fixed to an
adjustable height rig. However, the usage of active marker system is to provide the
initial position and heading information only, whereas the speed and position of the
gold club is measured using a XSENS Inertia Measurement Units (IMU). The
15
researcher aims to use the sensor fusion data to illustrate the feasibility of the motion
tracking setup. The developed hardware was tested for a full golf swing where the
maximum club speed recorded during the experiment was 6.5m/s upon ball contact.
Besides golf, researchers have also conducted experiments to implement the
usage of vision-based MOCAP to capture the motion of a badminton player. Salim et
al. (2010) used four Oqus cameras (high-speed camera) from Qualisys to record and
analyse the motion of players when performing badminton strokes. Experiments
were conducted involving student from UniMAP (Perlis) whom have basic
badminton knowledge and were between 22 to 28 years old. Figure 2.7 shows the
placement of markers on the player's arm and racket, eight markers were placed
along the arm. In this experiment, the researcher aims to investigate the difference
between forehand overhead smash performed by male and female in a mixed double
game. After a series of testing, Salim et al. (2010) concluded that a male player
achieves higher racket velocity than a female player. The male players were able to
reach racket velocity of 10.52 m/s, while female players reached racket velocity of 9
m/s. This is mainly due to the physiological difference between male and female.
Figure 2.7: The placement of markers on test subject (Salim et al., 2010)
Further research was conducted by Yang (2013) to investigate and analyse
the parameters related to forehand smash motion, where analysis for torso rotation
magnitude measurement is focused. In this research, video analysis was used to
obtain the coordinates of the human arm joint for a forehand smash stroke, from
where the torso rotational angle is calculated. Experiments were conducted using ten
Markers
16
badminton players, where their torso rotational and hand velocity data were collected
and analysed. Yang (2013) stated that of all other techniques, the forehand smash
uses the most power and is an important badminton stroke that is being used to win
points.
In another research done by Jiang & Wang (2013), in-depth analysis of the
smashing motion was further performed by identifying two different type of smashes
in badminton; which is the ground and in-air shuttle smash. Figure 2.8 and Figure 2.9
illustrates the smash motion captures for ground and in-air shuttle respectively. The
researcher used six QUALISYS-MCU500 high-speed infrared cameras to capture the
accurate 3D movement made by players when performing the smash. Hence, using
data from the MOCAP software, the velocity of the arm is calculated to determine
the arm joints stress level. Jiang & Wang (2013) stated that by obtaining this
information, the spiking action of players jumping to hit the shuttlecock is regulated.
Figure 2.8: Smash motion for ground shuttle (Jiang et al., 2013)
Figure 2.9: Smash motion for in-air shuttle (Jiang et al., 2013)
17
Analysis of the serve stroke has also been carried out by researchers
collectively, as it is one of the most important and basic stroke in the game of
Badminton (Ahmed et al., 2011; Loong Teng & Paramesran, 2011; Nagasawa et al.,
2012; Tsai et al., 2000). On the other hand, Ahmed et al. (2011) performed analysis
on the arm movements for forehand long and short serve using Canon Legria HF S10
high-speed cameras. For this, six male badminton player from Aligarh Muslim
University was required asked to perform forehand long and short service strokes a
few times repetitively. The recorded video footage was segmented and analysed
using the “Silicon Coach Pro 7” motion analysis software to extract biomechanical
variables. Ahmed et al. (2011) indicated that the elbow joint influenced the
shuttlecock velocity for the service stroke significantly.
The advantages of using vision-based MOCAP technology is noticeable in
the increasing number of studies using this technology. Current MOCAP devices that
are used by researchers are Kinect (Microsoft, 2010), Vicon Motion Camera (Vicon,
2004), Markerless Organic Motion (Organic Motion, 2006) and Qualisys Motion
Camera (Qualisys, 2004). These MOCAP technologies are widely being used to
reanimate the motions of a subject into 3D models because the post processing of
marker based MOCAP is smoother than videography. Although many research have
been conducted on analysing the movement and motion of a badminton player using
vision-based MOCAP technology, all of them have one disadvantage in common
which is an environmental limitation. The measurements are often performed in a
laboratory or studio setting, which does not comply well with the real competition or
training conditions (Yuan & Chen, 2014). Although, advancements in camera
technology might overcome this limitation, but then the operating cost of such
system would be higher. Furthermore, specific hardware and software are required to
process the data that also leads to higher operating cost.
2.2.2 Textile-Based Devices
Wearable technology refers to portable computers or electronic devices that are
incorporated into clothing or accessories which are comfortable to be worn on the
human body (Pantelopoulos & Bourbakis, 2010). Textile-based sensory devices are
referred to as embedded electronics that are integrated into wearables sports clothing
such as a wristband, knee guard, and elbow cap. This textile-based wearable
technology has seen significant advancement over the years where sensors can
18
provide sensory and scanning details such as biofeedback and tracking of
physiological functions (Salazar et al., 2010). Wearable sensors are mostly used to
monitor the human body’s physiological response to exercise and also to measure the
kinematic aspect of performance in some cases (Morris et al., 2009). Textile-based
sensors are used to monitor the movements of an athlete naturally with integrated
sensors attached to the parts of the body that need to be analysed.
Monitoring human motion using accelerometers attached to clothing has been
the fundamental design in many research that has been carried out to measure sports
movements (Bekdash et al., 2015; Fitzgerald et al., 2007; Pantelopoulos et al., 2010;
Salazar et al., 2010). Farringdon et al. (1999) demonstrates two methods to measure
human motion using textile-based sensors; first is by developing a sensor badge
system using two ADXL05 accelerometers stitched to a chest band to detect the
motion of the wearer. The developed sensor badge was successfully able to detect
only five different motion of the wearer mainly sitting, standing, lying, walking and
running. As a result, the researcher referred the first three movements as a simple
exercise, where fewer data samples were needed to differentiate the motion being
performed. However, the walking and running motion required a minimum of twenty
samples per second to be represented. Second, a sensor jacket which aims to detect
the posture and movement of the wearer using knitted conductive fabric. The
illustration of the knitted stretch sensor is shown in Figure 2.10, where the resistivity
increases when the fabric is stretched, and it is capable of being stretched over fifty
percent of its original length.
Figure 2.10: Knitted stretch sensor and the sensor jacket final design (Farringdon et
al., 1999)
Stretch
Sensor
Knitted conductive
tracking
Attached Stretch Sensor
19
The sensor strips are stitched on top of a jacket in eleven predetermined
positions mostly focusing on the shoulder and elbow joint as shown in Figure 2.10.
The analogue output of each sensor is converted using analogue to digital converter
because the output is used as input for the kinematic geometry model; where the
mobility of the wearer is mathematically modelled. Although the developed system
is capable of detecting human motion, further work is needed if this concept is to be
adapted for multimodal human-computer interfaces. In another study, Salazar et al.
(2010) expanded the idea of using sensor badge by Farringdon et al. (1999) by using
multiple sensor badges connected which are referred to as body sensor network
(BSN). The BSN system consists of two sensor badges that use a Freescale
MMA7260QT accelerometer and an InvenSense IDG-300 gyroscope; these sensors
are interconnected using conductive fabric.
The system was tested for usability in sports by conducting experiments,
where tests were performed by measuring the motion of a human walking, running,
playing tennis and swimming. Salazar et al. (2010) added the approach of using BSN
in signal monitoring is gaining a new foothold in sports performance analysis with
the numerous advances in integrated circuits (IC), wireless communications and
textile technology. Previous studies have primarily focused on using conductive
textile to integrate electronic sensors into wearable clothing. In particular, the usage
of accelerometer and gyroscope which require mechanical plug-ins to position the
sensor on the garment which is usually uncomfortable and complex (Shyr et al.,
2014). Therefore, an alternative method of using fabric-based sensor is proposed by
Jayaraman & Park (2005) as shown in Figure 2.11.
Figure 2.11: Novel fabric-based sensor design (US6970731 B1, 2005)
20
This concept eliminated the need to use adhesives and hard wiring to attach
the sensors to the human body to monitor vital signs and electrical impulses of the
wearer, as snap connector are used to transfer the electrical impulses received.
Although, the usage of these micromechanical sensors are required to measure the
angular acceleration and position. It is not appropriate for measuring biological
aspects of the human body like a heartbeat, breathing rate, sweat composition or
body fatigue level.
The number of research conducted using textile sensors has been increasing
over the decade because they are comfortable to wear, flexible and stretchable
(Carvalho et al., 2014, US6970731 B1, 2005; Shyr et al., 2014). Furthermore, they
can be used daily for continued health monitoring. Having the same objective,
Carvalho et al. (2014) conducted experiments by using fabric integrated electrodes to
monitor the electrocardiographic (ECG) and electromyography (EMG) signals.
Hence, Carvalho et al. (2014) used electrodes made of silver chloride electrode
(Ag/AgCl) covered in knitted multi filament polyamide yarn which has a low
resistance value and could withstand dimension modification. Figure 2.12 shows the
developed electrodes attached to a t-shirt, where snap connectors are used to extract
the electrical impulse when the sensor is in contact with the skin.
Figure 2.12: Textile with (a) embedded electrodes and (b) snap connector (Carvalho
et al., 2014)
The resistivity value of the sensor seen in Figure 2.12 (a) changes every time
the wearer breathes as this motion makes the knitted fabric sensor to expand and
contract. Carvalho et al. (2014) stated that one of the advantages of using fabric-
based sensors is the possibilities of reusing them with minimal maintenance. On the
other hand, Shyr et al. (2014) developed an experimental textile strain sensor. This
21
sensor consists of two electrodes and a substrate covered in elastic conductive
webbing, which works on the same principle as a textile sensor. The aim was to build
a wearable gesture-sensing device to monitor a player’s elbow and knee flexion
angles. Figure 2.13 shows the placement of the developed fabric strain sensor
attached to the player’s body using woven fabric, referred to as steady fabric.
Figure 2.13: The developed (a) fabric sensor for elbow (Shyr et al., 2014)
Based on the experimental results, Shyr et al. (2014) state that a good linear
relationship was obtained between the elastic conducting webbing resistance value
and the flexion angle during the stretch-recovery cycle. So far, all the studies that
have been discussed use the textile-based sensor to measure human motion, but they
are all limited to measuring only one parameter. Morris et al. (2009) and Coyle et al.
(2009) conducted a study on a full body system that can measure the physiological
signals and body kinematics of an athlete by using textile-based wearable sensors.
The developed whole body system is divided into three main sensor category. First,
measuring the body fluids mainly sweat. Sweating is a good sign, but an important
factor in athletic performance is rehydration (Morris et al., 2009). Therefore, an
athlete’s dehydration level is measured by the difference in their sweat pH value.
Hence, Coyle et al. (2009) developed a waistband that can monitor in real-
time the sweat activity produced by athletes as shown in Figure 2.14 (a). The pH
value of sweat is determined by analysing the colour change using LED-based
optical detection sensor. Coyle et al. (2009) mentioned that using the waistband to
obtain useful physiological information has an advantage of being non-invasive and
does not disturb the player's movement. Figure 2.14 (c) shows the experimental setup
to test the wearable pH sensor, where the data obtained from the pH sensor is
processed and transmitted wirelessly to a personal computer to display the sweat
composition in real-time. Bromocresol purple (BCP) dye is used to exhibit a colour
22
change from yellow to purple, depending on the sweat pH level. Therefore, the LED
optical detection is sampled every 0.2Hz to allows precise colour change detection.
Figure 2.14: (a) the waistband with (b) integrated pH sensor pocket and (c)
experimental set-up (Coyle et al., 2009)
Secondly, Morris et al. (2009) developed a chest band which is used to detect
and monitor an athletes breathing pattern. This band integrates into clothing worn by
athletes, where the data collected could indicate how well the athlete is handling
particular types of exercise. The breathing pattern is determined by measuring the
extension and contraction of the rib cage using stretchable Lycra PPy sensor, which
is fabric based conducting polymer polypyrrole (PPy) that changes conductivity
when stretched. The developed chest band with the sensor is shown in Figure 2.15,
where the system can detect breathing patterns of the athletes in real time.
Figure 2.15: Chest band to measure breathing rate (Morris et al., 2009)
(a)
(b)
(c)
23
Thirdly, Morris et al. (2009) proposed using fabric bend sensor to measure
joint flexion angles, especially human hand joints. The sensors were made by
attaching and stitching a few pieces of stretchable conductive fabric together. The
developed fabric bend sensor works on the same principle as a stretchable fabric
sensor, where the resistance value increases as it is bent or stretched (Carvalho et al.,
2014, US6970731 B1, 2005; Pantelopoulos et al., 2010; Shyr et al., 2014). Figure
2.16 shows a glove integrated with the developed bend sensor which is designed for
rehabilitation purposes, where it is used to detect patients finger movements after
experiencing a stroke (Morris et al., 2009). The author mentioned that this approach
could be developed further and be applied to evaluate any joint movements
depending on the required sports movements for various sports.
Figure 2.16: Full finger flexion (0), half extension (1) and full extension (2) (Morris
et al., 2009)
2.2.3 Flexion Based Monitoring System
A goniometer was the only available method to measure human joint angles during
dynamic movements before the introduction of vision and sensor-based measuring
device (Norkin & White, 2009). A goniometer is a type of instrument that measures
angles between two corresponding objects, where it can be used to measure the
flexion, abduction, extension and adduction of a human joint (Morrey, Askew, &
Chao, 1981). In the 1980’s the human joints were measured using a ruler goniometer
for many application such as medical and sports. Later in the 1990’s the usage of a
digital goniometer was implemented. However, it was not until the 2000’s where
measurements using sensors were taken.
Morrey et al. (1981) used a ruler goniometer to study the normal elbow
motion required to perform fifteen activities of daily living on thirty-three healthy
adults. The goniometer was attached to the subject's arm using velcro and was
24
required to perform the given task. However, one disadvantage was discovered
during testing; the goniometer was bulky since it had two rulers pivoted together at
the centre of an angle-measuring protractor. Therefore, researchers began searching
for other methods to perform this measurement and started experimenting with
stretchable fabric to measure joint angles (Gemperle et al., 1998, US6970731 B1,
2005). An example has been discussed previously, where Morris et al. (2009) used
stretchable fabric to measure stroke patients finger flexion.
However, advancements in sensor technology have enabled newer and more
advanced sensor with high precision capability to be developed; this contributed to
the development of flex sensors. Flex sensors are fabricated from film sensor
substrate which causes a mechanical stress on the conductive pattern that leads to
change in electrical resistance (Saggio et al., 2016). The flex sensor has widely been
used in many different types of research to control machines, artificial devices, and
physical activity measurements. More importantly, flex sensors have played a major
role in the development of many rehabilitation hardware for stroke patients
(Borghetti, Sardini, & Serpelloni, 2014). A general idea of how flex sensors are used
to measure finger flexion can be seen in the research done by Saggio (2011), which
investigates the total number of sensors needed to measure the human hand motion
successfully. Flex sensors have been used extensively in biomedical systems, and
Saggio (2011) aims to build a data glove for rehabilitation purpose. Figure 2.17
shows the developed data glove with flex sensors integrated into the glove surface.
Figure 2.17: Developed data glove (Saggio, 2011)
Saggio (2011) mentioned that the experiment results demonstrated good
performance in term of accuracy and repeatability of measures. The author also
REFERENCES
Ahmadi, A., Destelle, F., Monaghan, D., O’Connor, N. E., Moran, K. (2014). A
framework for comprehensive analysis of a swing in sports using low-cost
inertial sensors. In IEEE SENSORS 2014 Proceedings (Vol. 2014, pp. 2211–
2214). IEEE.
Ahmed, S., Hussain, I., Arshad Bari, M., Ahmad, A., Khan, A. (2011). Analysis of
Arm Movement in Badminton of Forehand Long and Short Service. Innovative
Systems Design and Engineering, 2(3), 13–17.
Ali, A. M. M., Ambar, R., Jamil, M. M. A., Wahi, A. J. M., & Salim, S. (2012).
Artificial Hand Gripper Controller via Smart Glove for Rehabilitation Process.
2012 International Conference on Biomedical Engineering (ICoBE),
(February), 300–304.
Ang, W. S., Chen, I.-M., & Yuan, Q. (2013). Ambulatory measurement of elbow
kinematics using inertial measurement units. 2013 IEEE/ASME International
Conference on Advanced Intelligent Mechatronics: Mechatronics for Human
Wellbeing, AIM 2013, 756–761.
Arduino. (2005a). Arduino - Arduino Board Uno. Retrieved June 6, 2016, from
https://www.arduino.cc/arduinouno
Arduino. (2005b). Arduino - Home. Retrieved August 12, 2016, from
https://www.arduino.cc/
Baca, A., & Kornfeind, P. (2006). Rapid Feedback Systems for Elite Sports Training.
IEEE Pervasive Computing, 5(4), 70–76.
Bartlett, R. (2007). Introduction to Sports Biomechanics. Sports Biomechanics (2nd
Editio). Abingdon, UK: Taylor & Francis.
Bekdash, O., Norcross, J., Science, W., Group, E., Engineer, P. (2015). Monitoring
Human Performance during Suited Operations : A Technology Feasibility Study
Using EMU Gloves. 45th International Conference on Environmental Systems,
Washington, (July), 1–11.
149
Boga, S. (2008). Badminton. Paw Prints.
Borghetti, M., Sardini, E., & Serpelloni, M. (2014). Evaluation of bend sensors for
limb motion monitoring. IEEE MeMeA 2014 - IEEE International Symposium
on Medical Measurements and Applications, Proceedings, 1–5.
Buttimer, E., & Luttrell, S. (2010). Badminton - Scheme Of Work.
Carvalho, H., Catarino, A. P., Rocha, A., & Postolache, O. (2014). Health monitoring
using textile sensors and electrodes: An overview and integration of
technologies. In 2014 IEEE International Symposium on Medical Measurements
and Applications (MeMeA) (pp. 1–6). IEEE.
Chang, K.-H., & Consulting, C. (2014). Bluetooth: a viable solution for IoT [Industry
Perspectives]. IEEE Wireless Communications, 21(6), 6–7.
Cisco. (2015). Internet of Things (IoT). Retrieved June 6, 2016, from
http://www.cisco.com/c/en/us/solutions/internet-of-things/overview.html
Cobos, S., Ferre, M., & Aracil, R. (2010). Simplified human hand models based on
grasping analysis. In 2010 IEEE/RSJ International Conference on Intelligent
Robots and Systems (pp. 610–615). IEEE.
Cobos, S., Ferre, M., Sanchez Uran, M. A., Ortego, J., & Pena, C. (2008). Efficient
human hand kinematics for manipulation tasks. In 2008 IEEE/RSJ International
Conference on Intelligent Robots and Systems (pp. 2246–2251). IEEE.
Coyle, S., Morris, D., Lau, K.-T., Diamond, D., Luprano, J. (2009). Textile sensors
to measure sweat pH and sweat-rate during exercise. In Proceedings of the 3d
International ICST Conference on Pervasive Computing Technologies for
Healthcare (pp. 4–9). ICST.
Cutti, A. G., Giovanardi, A., Rocchi, L., Davalli, A., & Sacchetti, R. (2008).
Ambulatory measurement of shoulder and elbow kinematics through inertial
and magnetic sensors. Medical and Biological Engineering and Computing,
46(2), 169–178.
Farringdon, J., Moore, A. J., Tilbury, N., Church, J., & Biemond, P. D. (1999).
Wearable sensor badge and sensor jacket for context awareness. In Digest of
Papers. Third International Symposium on Wearable Computers (pp. 107–113).
IEEE Comput. Soc.
Federation, B. W. (2006). Badminton History. Retrieved June 6, 2016, from
http://bwfbadminton.org/page.aspx?id=14887
Fischer, R. (2013). History of Motion Capture. Retrieved September 5, 2016, from
150
http://www.motioncapturesociety.com/resources/industry-history
Fitzgerald, D., Foody, J., Kelly, D., Ward, T., Caulfield, B. (2007). Development of a
wearable motion capture suit and virtual reality biofeedback system for the
instruction and analysis of sports rehabilitation exercises. Annual International
Conference of the IEEE Engineering in Medicine and Biology - Proceedings,
4870–4874.
Fong, D. T. W., Wong, J. C. Y., Lam, A. H. F., Lam, R. H. W., & Li, W. J. (2004). A
wireless motion sensing system using ADXL MEMS accelerometers for sports
science applications. In Fifth World Congress on Intelligent Control and
Automation (Vol. 6, pp. 5635–5640). Hangzhou: IEEE.
Gaston, R. G., & Loeffler, B. J. (2015). Sports-Specific Injuries of the Hand and
Wrist. Clinics in Sports Medicine.
Gemperle, F., Kasabach, C., Stivoric, J., Bauer, M., & Martin, R. (1998). Design for
wearability. Digest of Papers. Second International Symposium on Wearable
Computers, 116–122.
Generation, N. (1995). Motion Capture: Optical Systems. Imagine Media, (10), 53.
Grice, T. (2008). Badminton : Steps to Success (2nd ed.). Human Kinetics.
Group, S. I. (2014). Bluetooth Technology Website. Retrieved April 10, 2015, from
http://www.bluetooth.com/Pages/Bluetooth-Home.aspx
Guillain, J.-Y. (2004). Badminton: An Illustrated History. Publibook.
Guinness World Records. (2015). Fastest Badminton hit in competition (male).
Retrieved from http://www.guinnessworldrecords.com/world-records/fastest-
badminton-hit-in-competition-(male)
Healthline, M. (2005). The Human Arm. Retrieved from
http://www.healthline.com/human-body-maps/arm#seoBlock
Henao, Y. H., & Fruett, F. (2009). Development of an electromechanical sensor
system to monitor sports activities. In 2009 9th International Conference on
Electronic Measurement & Instruments (Vol. 4, pp. 4-6-4–10). IEEE.
Hiki, S., Ikehara, W., Murayama, K., Imaizumi, K., & Fukuda, Y. (1998).
Description of movement of the upper extremities based on anatomical structure
model (with special regard to signing gesture). In Proceedings of the 20th
Annual International Conference of the IEEE Engineering in Medicine and
Biology Society. Vol.20 Biomedical Engineering Towards the Year 2000 and
Beyond (Vol. 5, pp. 2444–2447). IEEE.
151
Ho, D. (2003). Notepad ++. Retrieved September 22, 2016, from https://notepad-
plus-plus.org/
Hong, E. (2005). Hand injuries in sports medicine. Prim Care, 32(1), 91–103.
Hopley, M. (2005). Badminton Bible. Retrieved December 16, 2015, from
https://www.badmintonbible.com/
Information, B. (2006). Badminton Grip. Retrieved June 20, 2015, from
http://www.badminton-information.com/badminton_grip.html
InvenSense. (2003). MPU-6050. InvenSence.
ITeadStudio. (2014). HC-05 - Serial Port Bluetooth Module (Master/Slave).
Retrieved April 5, 2016, from http://wiki.iteadstudio.com/Serial_Port_Bluetooth
_Module_(Master/Slave)_:_HC-05
Jaitner, T., & Gawin, W. (2007). Analysis of Badminton Smash With a Mobile
Measure Device Based on Accelerometry. XXV ISBS Symposium, Ouro Preto,
Brazil, 282–284.
Jaitner, T., & Gawin, W. (2010). A mobile measure device for the analysis of highly
dynamic movement techniques. Procedia Engineering, 2(2), 3005–3010.
Jayaraman, S., & Park, S. (2005). US6970731 B1. Google Patents.
Jiang, R., & Wang, Z. (2013). Analysis on Smashing Motion in Badminton. In
Lecture Notes in Electrical Engineering (Vol. 217 LNEE, pp. 651–657).
Johnson, L. (2011). Active Markers. Retrieved September 5, 2016, from
https://www.cs.utexas.edu/~dana/vrlab/research.html
Jones, B., & Jarvis, H. (1993). Badminton Royal Navy. (P. Edwards, Ed.) (1st ed.).
London: Education and Youth Limited.
Jung, P.-G., Lim, G., & Kong, K. (2012). Human posture measurement in a three-
dimensional space based on inertial sensors. Control, Automation and Systems
(ICCAS), 2012 12th International Conference on, 1013–1016.
Kenn, L. (2014). Kuler Badminton Grips. Retrieved August 23, 2016, from
http://kulergrip.com/blog/4-basic-badminton-grip.html
Kiang, C. T., Yoong, C. K., & Spowage, A. C. (2009). Local sensor system for
badminton smash analysis. In 2009 IEEE Intrumentation and Measurement
Technology Conference, I2MTC 2009 (pp. 889–894). IEEE.
King, K., Yoon, S. W., Perkins, N. C., & Najafi, K. (2008). Wireless MEMS inertial
sensor system for golf swing dynamics. Sensors and Actuators, A: Physical,
141(2), 619–630.
152
Kortier, H., Sluiter, V., Roetenberg, D., & Veltink, P. (2014). Assessment of hand
kinematics using inertial and magnetic sensors. Journal of NeuroEngineering
and Rehabilitation, 11(1), 70.
Lin, Z.-M., Chang, C., Chou, N., & Lin, Y. (2014). Bluetooth Low Energy (BLE)
based blood pressure monitoring system. In 2014 International Conference on
Intelligent Green Building and Smart Grid (IGBSG) (pp. 1–4). IEEE.
Liu, T., Inoue, Y., & Shibata, K. (2006). A wearable sensor system for human
motion analysis and humanoid robot control. 2006 IEEE International
Conference on Robotics and Biomimetics, ROBIO 2006, 43–48.
Loong Teng, S., & Paramesran, R. (2011). Detection of service activity in a
badminton game. In TENCON 2011 - 2011 IEEE Region 10 Conference (pp.
312–315). IEEE.
Lu, Z., Su, B., Wang, G., Cao, H., Ge, Y. (2013). Sports biomechanics infomation
acquisition and analysis based on Miniature Inertial Measurement. In 2013
IEEE International Conference on Cyber Technology in Automation, Control
and Intelligent Systems (pp. 144–148). IEEE.
Luinge, H. J., Veltink, P. H., & Baten, C. T. M. (2007). Ambulatory measurement of
arm orientation. Journal of Biomechanics, 40(1), 78–85.
MacKenzie, B. (1997). Performance Evaluation Tests. Retrieved August 21, 2015,
from http://www.brianmac.co.uk/eval.htm
Mertz, L. (2013). Technology Comes to the Playing Field: New World of Sports
Promises Fewer Injuries, Better Performance. IEEE Pulse, 4(5), 12–17.
Microsoft. (2002). Microsoft Visual Studio. Retrieved September 20, 2016, from
http://visualstudio.com/
Microsoft. (2010). Kinect. Retrieved from https://en.wikipedia.org/wiki/Kinect
Mirabella, O., Raucea, A., Fisichella, F., & Gentile, L. (2011). A motion capture
system for sport training and rehabilitation. 4th International Conference on
Human System Interaction, HSI 2011, 52–59.
Moeslund, T. B., & Granum, E. (2001). A Survey of Computer Vision-Based Human
Motion Capture. Computer Vision and Image Understanding, 81(3), 231–268.
Moeslund, T. B., Hilton, A., & Krüger, V. (2006). A survey of advances in vision-
based human motion capture and analysis. Computer Vision and Image
Understanding, 104(2–3), 90–126.
Mohan, A., Devasahayam, S. R., Tharion, G., & George, J. (2013). A Sensorized
153
Glove and Ball for Monitoring Hand Rehabilitation Therapy in Stroke Patients.
India Educators’ Conference (TIIEC), 2013 Texas Instruments, 321–327.
Montazerian, R., Shaheen, A., Eftaxiopoulou, T., & Bull, A. M. (2008). 3-D Elbow
Kinematics During Cricket Bowling. Journal of Biomechanics, 41(July), S85.
Morrey, B. F., Askew, L. J., & Chao, E. Y. (1981). A biomechanical study of normal
functional elbow motion. The Journal of Bone and Joint Surgery. American
Volume, 63(6), 872–877.
Morris, D., Coyle, S., Lau, K.-T., Diamond, D., & Moyna, N. (2009). Textile-Based
Wearable Sensors for Assisting Sports Performance. 2009 Sixth International
Workshop on Wearable and Implantable Body Sensor Networks, 307–311.
Morrow, J. R., Jackson, A. W., & Disch, J. G. (2011). Measurement and Evaluation
in Human Performance (4th ed.). book, Human Kinetics.
Morrow, J. R., Jackson, A. W., Disch, J. G., & Mood, D. P. (2005). Measurement
and Evaluation in Human Performance (3rd ed.). Human Kinetics.
Motion Workshop. (2007). MotionNode. Retrieved July 12, 2016, from
http://www.motionnode.com/
Nagasawa, M., Hatori, Y., Kakuta, M., Hayashi, T., & Sekine, Y. (2012). Smash
motion analysis for badminton from image. Proceedings of the IEEJ Image
Electronics and Visual Computing Workshop, 1–8.
Nam, C. N. K., Kang, H. J., & Suh, Y. S. (2014). Golf swing motion tracking using
inertial sensors and a stereo camera. IEEE Transactions on Instrumentation and
Measurement, 63(4), 943–952.
Noiumkar, S., & Tirakoat, S. (2013). Use of optical motion capture in sports science:
A case study of golf swing. Proceedings - 2013 International Conference on
Informatics and Creative Multimedia, ICICM 2013, 310–313.
Norkin, C. C., & White, D. J. (2009). Measurement of Joint Motion: A Guide to
Goniometry. F.A. Davis.
O’Donoghue, P. (2009). Research Methods for Sports Performance Analysis.
(Routledge, Ed.). Taylor & Francis.
OptiTrack. (1996). OptiTrack. Retrieved January 3, 2016, from http://optitrack.com/
Oracle. (2004). About Java and Java +. Retrieved January 5, 2017, from
https://www.java.com/en/
Organic Motion. (2006). Organic Motion Motion Capture. Retrieved from
http://www.organicmotion.com/
154
Orthopaedic Surgeons, A. A. of. (1965). Joint Motion: Method of Measuring and
Recording. (P. S. University, Ed.). The Academy.
Pantelopoulos, A., & Bourbakis, N. G. (2010). A survey on wearable sensor-based
systems for health monitoring and prognosis. IEEE Transactions on Systems,
Man and Cybernetics Part C: Applications and Reviews, 40(1), 1–12.
Papas, M. (2008). Wrist use. Retrieved December 16, 2015, from
http://www.revolutionarytennis.com/wristuse.html
Patel, S., Park, H., Bonato, P., Chan, L., & Rodgers, M. (2012). A review of
wearable sensors and systems with application in rehabilitation. Journal of
NeuroEngineering and Rehabilitation, 9(1), 21.
Qualisys. (2004). Qualisys Motion Capture. Retrieved December 17, 2016, from
www.qualisys.com/applications/sports
Rhodes, B., & Mase, K. (2006). Wearables in 2005. IEEE Pervasive Computing,
5(1), 92–95.
Rocha, L. A., & Correia, J. H. (2006). Wearable Sensor Network for Body
Kinematics Monitoring. In 2006 10th IEEE International Symposium on
Wearable Computers (pp. 137–138). IEEE.
Rowberg, J. (2012). MPU-6050 6-axis accelerometer/gyroscope. Retrieved
September 22, 2016, from http://www.i2cdevlib.com/devices/mpu6050#source
Rusydi, M. I., Sasaki, M., Sucipto, M. H., Zaini, & Windasari, N. (2015a). Local
Euler Angle Pattern Recognition for Smash and Backhand in Badminton Based
on Arm Position. Procedia Manufacturing, 3(Ahfe), 898–903.
Rusydi, M. I., Sasaki, M., Sucipto, M. H., Zaini, & Windasari, N. (2015b). Study
about backhand short serve in badminton based on the Euler angle. In 2015 4th
International Conference on Instrumentation, Communications, Information
Technology, and Biomedical Engineering (ICICI-BME) (pp. 108–112). IEEE.
Saggio, G. (2011). Bend sensor arrays for hand movement tracking in biomedical
systems. Proceedings of the 4th IEEE International Workshop on Advances in
Sensors and Interfaces, IWASI 2011, 51–54.
Saggio, G., Riillo, F., Sbernini, L., & Quitadamo, L. R. (2016). Resistive flex
sensors: a survey. Smart Materials and Structures, 25(1), 13001.
Salazar, A. J., Silva, A. S., Borges, C. M., & Correia, M. V. (2010). An initial
experience in wearable monitoring sport systems. In Proceedings of the 10th
IEEE International Conference on Information Technology and Applications in
155
Biomedicine (pp. 1–4). IEEE.
Salim, M. S., Lim, H. N., Salim, M. S. M., & Baharuddin, M. Y. (2010). Motion
analysis of arm movement during badminton smash. In Proceedings of 2010
IEEE EMBS Conference on Biomedical Engineering and Sciences, IECBES
2010 (pp. 111–114). IEEE.
Sazali. (2015). Personal Communication, Interview Method
Senanayake, S. M. N. A. (2004). Walking, running and kicking using body-mounted
sensors. In IEEE Conference on Robotics, Automation and Mechatronics, 2004.
(Vol. 2, pp. 1141–1146). IEEE.
Shan, C. Z., Ming, E. S. L., Rahman, H. A., & Yeong Che Fai. (2015). Investigation
of upper limb movement during badminton smash. In 2015 10th Asian Control
Conference (ASCC) (pp. 1–6). IEEE.
Shan, C. Z., Sen, S. L., Che Fai, Y., & Su Lee Ming, E. (2015). Investigation of
Sensor-based Quantitative Model for Badminton Skill Analysis and
Assessment. Jurnal Teknologi, 72(2), 93–96.
Shingade, A., & Ghotkar, A. (2014). Animation of 3D Human Model Using
Markerless Motion Capture Applied To Sports. International Journal of
Computer Graphics & Animation, 4(1), 27–39.
Shirota, K., Watanabe, K., & Kurihara, Y. (2012). Measurement and analysis of golf
swing using 3-D acceleration and gyro sensor. Proceedings of the Society of
Instrument and Control Engineers Conference, 356–360.
Shyr, T.-W., Shie, J.-W., Jiang, C.-H., & Li, J.-J. (2014). A Textile-Based Wearable
Sensing Device Designed for Monitoring the Flexion Angle of Elbow and Knee
Movements. Sensors, 14(3), 4050–4059.
Sirio, G. Di. (2006). ChibiOS. Retrieved October 10, 2016, from
http://www.chibios.org/dokuwiki/doku.php
Spectrasymbols. (2014). Flex Sensor: How It Works (Vol. 1).
Takeda, R., Tadano, S., Natorigawa, A., Todoh, M., & Yoshinari, S. (2009). Gait
posture estimation using wearable acceleration and gyro sensors. Journal of
Biomechanics, 42(15), 2486–2494.
Taniguchi, A., Watanabe, K., Kurihara, Y., & Sice. (2012). Measurement and
analyze of jump shoot motion in basketball using a 3-D acceleration and
gyroscopic sensor. 2012 Proceedings of Sice Annual Conference (Sice), 361–
365.
156
Taylor, L., Miller, E., & Kaufman, K. R. (2017). Static and dynamic validation of
inertial measurement units. Gait & Posture, 57(April), 80–84.
Tong, Y.-M., & Hong, Y. (2000). The playing pattern of world’s top single
badminton players. 18 International Symposium on Biomechanics in Sports, 1–
6.
Tran, N. (2010). Badminton : Proper Racket Grip in Badminton. University of
California, United States: YouTube.
Tsai, C.-L., Huang, C., Lin, D.-C. C., Cheng, C.-C., Chang, S. S. (2000).
Biomechanical analysis of the upper extremity in three different badminton
overhead strokes. In ISBS-Conference Proceedings Archive (Vol. 1, pp. 35–38).
Turner, C. (2001). NiCd and NiMh batteries. Retrieved September 21, 2016, from
http://www.ka7oei.com/nicds.html
Vicon. (2004). Vicon Motion Capture. Retrieved from https://www.vicon.com/vicon
Virgala, I., Kelemen, M., & Mrkva, Š. (2013). Kinematic Analysis of Humanoid
Robot Hand. American Journal of Mechanical Engineering, 1(7), 443–446.
Wang, Y., Chang, S., Shan, G., & Li, H. (2014). A Wireless Sensor System for the
Training of Hammer Throwers. In 2014 Tenth International Conference on
Computational Intelligence and Security (pp. 620–623). IEEE.
Wang, Z., Guo, M., & Zhao, C. (2016). Badminton Stroke Recognition Based on
Body Sensor Networks. IEEE Transactions on Human-Machine Systems, 73(4),
1–7.
Weinberg, H. (2010). The Five Motion Senses : Using MEMS Inertial Sensing to
Transform Applications. Sensors, (Analog Devices), 1–5.
Wong, C. (2012). Badminton - Basic Techniques. Retrieved August 21, 2016, from
http://badminton.chorwong.com/badmintontechniques.html
Xsens. (2009). MTi. Retrieved from https://www.xsens.com/products/mti/
Xu, J. Y., Nan, X., Ebken, V., Wang, Y., Kaiser, W. J. (2015). Integrated Inertial
Sensors and Mobile Computing for Real-Time Cycling Performance Guidance
via Pedaling Profile Classification. IEEE Journal of Biomedical and Health
Informatics, 19(2), 440–445.
Yang, N. (2013). Research of badminton forehand smash technology based on
biomechanical analysis. Journal of Chemical and Pharmaceutical Research,
5(11), 172–177.
Yap, C. (2006). Badminton Information. Retrieved August 12, 2016, from
157
http://www.badminton-information.com/
Yuan, Q., & Chen, I.-M. (2014). Localization and velocity tracking of human via 3
IMU sensors. Sensors and Actuators A: Physical, 212, 25–33.
Zatsiorsky, V. M. (1998). Kinematics of Human Motion. (H. Kinematic, Ed.) (1st
ed.). Human Kinetics.