Impact Distance Detection in Tennis Forehand by an Inertial System

Lucio Caprioli

1 a

, Francesca Campoli

1 b

, Saeid Edriss

1 c

, Elvira Padua

2 d

, Amani Najlaoui

1 e

Cristian Romagnoli

2 f

, Giuseppe Annino

3 g

and Vincenzo Bonaiuto

1 h

Sports Engineering Lab., Dept. Industrial Eng. Univ. Rome Tor Vergata, Rome, Italy

Dept. of Human Science & Prom. of Quality of Life, San Raffaele Rome Univ., Rome, Italy

Dept. of Medicine Systems, Univ. Rome Tor Vergata, Rome, Italy

Keywords: Inertial Measurement System, Sports Performance Assessment, Tennis, Forehand, Lateral Distance.

Abstract: Assessing the distance to the ball in the tennis forehand is fundamental. In this context, a non-invasive

assessment system can help technicians even more in amateur tennis, where players who are still unaware of

the act need continuous feedback. Three amateur tennis players with an average of 4 years of playing

experience were recruited. The subjects wore a sensorized chest strap with an inertial unit and received two

sets of 10 balls each. Two action cameras captured 20 forehands of each player from lateral and rear

perspectives, aligned about 6m from the point of impact. Video analysis was conducted to identify the

anteroposterior and lateral distance of the ball at the point of impact from the longitudinal axis coincident

with the first toe of the nondominant foot. Pearson's correlation between distance and trunk inclination during

the impact phase was investigated, and a strong correlation was found for all the subjects. This prompts us to

consider the potential of a sensorized chest strap to assess the individual optimal distance from the ball in the

forehand of tennis amateurs. Subsequent studies are needed to develop the system's full potential, expand the

number of subjects, and examine all the fundamentals of the game.

1 INTRODUCTION

Tennis is a highly technical sport and requires fine

motor coordination (Casale, 2003; Roetert & Kovacs,

2019). The learning process is particularly

demanding, and it is common to make various

mistakes in improving skills (Reid et al., 2013). The

technique can be continuously refined (Castellani et

al., 2007), and biomechanical analysis of the

movement can effectively correct improper actions.

Studies comparing high-level and amateur players

have provided insights into optimal angles and body

positions during various strokes (Fleisig et al., 2003;

Knudson & Elliott, 2004; Landlinger et al., 2010;

Nesbit et al., 2008; Reid & Elliott, 2002; Roetert &

https://orcid.org/0009-0005-4049-5225

https://orcid.org/0009-0004-1342-5881

https://orcid.org/0009-0000-0224-8294

https://orcid.org/0000-0001-5227-2567

https://orcid.org/0009-0001-0153-7251

https://orcid.org/0000-0003-0904-634X

https://orcid.org/0000-0001-8578-6046

https://orcid.org/0000-0002-2328-4793

Kovacs, 2019). However, playing technique is

strongly influenced by the tactical context and how

the player reaches the position. That is why, in recent

years, the concept of pure technique has lost its value,

and coaches are increasingly talking about technique

applied to the tactical context (Castellani et al., 2007).

Even in simple, controlled situations in training,

amateur players have difficulty always positioning

themselves at an optimal distance from the ball,

presenting a higher variability in stroke execution,

unlike advanced athletes (Caprioli et al., 2024). This

variability occurs because less experienced players

are less able to read the ball's trajectory and less aware

of their technical gestures.

For this reason, assessing the distance to the ball

in rebound shots (forehand and backhand) is critical,

Caprioli, L., Campoli, F., Edriss, S., Padua, E., Najlaoui, A., Romagnoli, C., Annino, G. and Bonaiuto, V.

Impact Distance Detection in Tennis Forehand by an Inertial System.

DOI: 10.5220/0013073100003828

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 12th International Conference on Sport Sciences Research and Technology Support (icSPORTS 2024), pages 277-282

ISBN: 978-989-758-719-1; ISSN: 2184-3201

277

particularly for amateur tennis players. Proper

distance maximizes ball contact in the ideal area of

the racquet, known as the "sweet spot" allowing more

control over the direction and power of the shot. If the

impact occurs too close to the body the extension of

the arm is limited, and as well as in the case of

excessive distance reduced power and accuracy of the

shot. Keeping the right distance from the ball can also

help reduce the risk of injury, as it allows you to

execute the shot with proper technique and without

excessive stress on joints, muscles, and ligaments (Fu

et al., 2018; Reid et al., 2013).

1.1 Forehand Shot Assessment

Forehand is considered the most important technical

gesture after the serve (Johnson & McHugh, 2006),

and in the game, forehand shots are about 25 percent

more than backhand shots. Players are constantly

striving to improve their forehand from a technical

standpoint. In the forehand, we can observe different

personalisms among tennis players; however, some

technical parameters are present in all of them, and it

can be divided into six main phases: starting position,

preparation (unit-turn), opening phase, stance,

impact, final (Bertino et al., 2012). In the forehand,

the greatest linear and internal rotation velocity

expressions occurred quite late in the forward swing

phase toward impact (Elliott, 2006; Elliott et al.,

1997; Seeley et al., 2016).

Among the methods currently most widely used

for technical assessment are 2D and 3D video motion

tracking systems (Annino et al., 2023; Edriss et al.,

2024; Lambrich & Muehlbauer, 2023; Martin et al.,

2021). However, these systems often prove to be

expensive or time-consuming. For this purpose,

proposed as a valid alternative solution to optical

motion tracking systems the inertial sensors (Inertial

Motion Unit – IMU) (Delgado-García et al., 2021;

Hernández-Belmonte & Sánchez-Pay, 2021; Zanela

et al., 2024) IMUs contain MEMS-type

accelerometers, gyroscopes, and magnetometers, and

through a sensor fusion estimation 3D orientation is

obtained. The advantages of inertial sensors include,

among others, the accuracy, convenience, and

quickness of measuring and analyzing data, which

can potentially take place even in real-time without

the need for very expensive hardware components,

unlike current 3D motion capture systems (Edriss et

al., 2024). Despite their small size and applicability in

almost any environment, these devices face several

technical challenges (Alcala et al., 2021). In some

studies, inertial sensors were placed on the racket or

wristband for stroke analysis and classification

(Ebner & Findling, 2019; Kos & Kramberger, 2018).

Although this approach allows the direct

measurement of important kinematic parameters such

as accelerations, angular velocities, and the exact

position of joint segments using quaternions, the

extensive use of these devices may prove invasive.

Although very lightweight, several sensors placed on

the tennis player's body could sensitively limit

movement or otherwise impair the naturalness of the

gesture. In the same way, even a single sensor of a

few grams placed on the tennis racket can

compromise the technique of the strokes.

The help a non-invasive assessment system can

provide a technician is even more apparent in amateur

tennis, where players still unaware of the act need

continuous feedback. In this study, an inertial

measurement system was applied to detect the

distance from the ball at the impact in the forehand

shot of 3 amateur players.

2 MATERIALS AND METHODS

The analysis was conducted on three amateur tennis

players (one female and two males; 32.7 ± 6.8 years,

175.3 ± 8.2 cm) with an average of 4 years of playing

experience. All recruited subjects were in good

health, had not suffered recent injuries, and consented

to data processing for research purposes. The subjects

wore a sensorized chest strap on which a Movella

DOT inertial unit was mounted (Table 1). The device

was positioned as tightly as possible without allowing

the sensor to move freely and as comfortably as

possible for each subject, depending on the physical

conformations of each male and female. The inertial

sensors, previously calibrated, were set to an offline

acquisition mode at 120 Hz.

Table 1: Inertial Measurement Unit details.

Dimensions:

36.3 × 30.35 × 10.8 mm

Weight:

11.2 g

Recording Mode

Offline

Sampling Rate

120Hz

Connection

Bluetooth 5.0

All measurements were conducted on a sunny day

with no adverse wind conditions in an outdoor court.

The subjects followed a 20-minute physical and

technical warm-up and then played two sets of 10

forehand strokes in a controlled situation, with a

three-minute rest between sets. A Tennis Tutor Plus

ball-launching machine, positioned on the opposite

baseline at 1.60m from the mid-point, was set to two

settings for the two practice sets. For the first set,

icSPORTS 2024 - 12th International Conference on Sport Sciences Research and Technology Support

278

subjects received ten balls each, at speed 4, with no

effect and easy to handle. In the second, ten balls at

speed 6 with top spin instead. Two action cameras

240 Hz captured 20 forehands of each player from

lateral and rear perspectives, both aligned about 6m

from the point of impact and placed on a stand at

1.10m above the ground. The cameras were

synchronized with each other through a luminous

pulse. Video analysis was conducted using Kinovea

software (version 0.9.5) (Charmant & contributors,

2021) to identify the lateral and anteroposterior

distance of the ball at the point of impact concerning

the longitudinal axis coincident with the first toe of

the nondominant foot (Figure 1). Since the players did

not have the same position on the court in each shot,

in each impact frame, calibration was re-performed

using the racquet of a known size equivalent to 68.50

cm as a reference placed on the measurement plane

(subjects used standard-length racquet models).

Figure 1: Distance measurement by video analysis. The

racket measurement was used as a reference placed on the

measurement plane in the calibration process.

2.1 Data Analysis

From the data acquired by the IMU, the trunk

inclination angle (Euler Y) during the impact phase,

coinciding with the point of the maximum angular

velocity of trunk rotation (Gyr X), was analyzed. The

Shapiro-Wilk test was used to validate the assumption

of normality. Since no significant deviations from

normality were detected, parametric tests were used

for inference. The coefficient of variation (CV) for

repeated measurement, interclass correlation

coefficients (ICC), standard error of measurement

(SEM) and 95% confidence interval (95% CI) were

calculated to determine the set-to-set reliability for

lateral and anteroposterior distance, and trunk angle

(Euler Y). Moreover, the ICC was used as an

assessment test of consistency and the repeatability of

quantitative measurements made by the same

operator in two different sets. Paired t-tests and the

Pearson correlation coefficient (r) were used for

repeatability of test-re-test measurements. In

addition, the effect sizes (ES) were also calculated

using Cohen's d between the first set and the second

set of means (Cohen, 2013), where the small effect

was 0.1, moderate 0.3, and large was 0.5 (Cooper et

al., 2019).

Pearson's correlation between ball distance and

trunk inclination during the impact was investigated

for each player and the group. MedCalc software

(Version 23.0.2) was used for statistical analysis.

3 RESULTS

3.1 Reliability

Test-retest values of Mean, SD, SEM, ICC, Pearson

correlation coefficient (r), and the CV relative to the

lateral distance (LD), trunk inclination (Euler Y), and

anteroposterior distance (APD) performed in the two

sets are reported in Table 2.

Table 2: Set-to-set repeatability of average lateral distance LD (cm) and trunk angles Euler Y of the forehands performed by

three amateur tennis players. r Pearson correlation coefficient; CV, Coefficient of Variation for repeated measurements; ICC,

Interclass Correlation Coefficient; 95% Confidence Interval (CI); SEM, Standard Error of Measurement; and ES, Effect Size.

Different

Set

Set 1

Set 2

ICC

95% CI

SEM

Parameters

Mean ± SD

LD (cm)

88.43 ± 16

87.45 ± 23

0.998

4.75

0.966

-0,3021 to 0,9991

0,730

0,382

Euler Y (°)

75.50 ± 5

77.03 ± 4

0.902

2.13

0.934

-1,5807 to 0,9983

1,052

-0,049

APD (cm)

-6,77 ± 18

-12,42 ± 23

0.972

244.54

0,968

-0,2674 to 0,9992

2,881

0,272

Root mean square method

Impact Distance Detection in Tennis Forehand by an Inertial System

279

ICC found no significant differences between the

first and second sets of measurements, showing high

reproducibility for all measurements. The CV was

high or extremely high in the case of APD. The effect

size is small or moderate in the case of LD. The two

measurements have a strong Pearson correlation

(Table 2).

3.2 Descriptive Statistics

Sixty forehand shots were analyzed. The mean lateral

distance measured among all trials was 88.52cm with

a standard deviation of ± 18cm, with a minimum

value of 61cm and a maximum value of 120cm. The

mean anteroposterior distance was -9.42cm ± 23cm

with a minimum value of -53cm and a maximum

value of 20cm. As for trunk inclination (Euler Y), it

was an average of 76.30° with a standard deviation of

5°, minimum value of 63° and maximum value of

85°.

3.3 Inferential Statistics

A highly significant strong Pearson's correlation was

found in all the subjects between trunk angle Euler Y

(i.e., flexion-extension angle) and lateral distance

from the ball. In particular, in Player 1 was found a

medium-high correlation (r= 0.69 p<0.001) (Figure

2), strong in Player 2 (r= 0.79 p<0.001) and Player 3

(r= 0.70 p<0.001) (Figure 3). A more moderate partial

correlation (r= 0.40 p<0.001) was found in the

analysis of the 60 forehand shots of the whole group

due to the variability of distance to the ball in players

with different body stature and joint levers. However,

no significant correlations were found with

anteroposterior distance.

Figure 2: Pearson's correlation in Player 1.

(a)

(b)

Figure 3: Pearson's correlation in Player 2 (a) and 3 (b).

4 DISCUSSION

This study allowed us to evaluate, even if in

preliminary form and on a sample of only three

subjects, the reliability of a system for distance

assessment in the forehand based on a single inertial

sensor applied on the athlete through a chest strap.

The measurements were reliable, and a strong,

significant correlation was found between torso tilt

and lateral distance. The correlation found confirms

what was logically hypothesized at the beginning of

the study: a proper distance from the ball may allow

for better body weight transfer through forward torso

tilt, as opposed to distances that are too short. It

should be considered from a simple observation that

icSPORTS 2024 - 12th International Conference on Sport Sciences Research and Technology Support

280

amateur players generally tend not to position

themselves at a sufficient distance from the ball. In

the face of these results, the study opens new on-court

applications for improving amateur tennis

performance through personalized feedback. Indeed,

it is possible to develop, via the SDK provided by the

manufacturer, simple mobile apps connected via

Bluetooth to the sensor that can indicate a customized

correction in real-time. The study's main limitation is

the small number of subjects, which will need to be

expanded. Furthermore, in this study, only the

forehand technique was examined without analyzing

the stroke result, and because of the need to

standardize the investigation protocol, a single

structured situation was assessed, and not all possible

game situations that may occur during a match were

examined.

5 CONCLUSIONS

This preliminary study found a strong correlation

between the torso tilt detected by the IMU system and

the lateral distance to the ball at the impact point. This

bodes well for how a sensorized chest strap can aid

the technician in assessing the individual optimal

distance to the ball in the forehand of amateur tennis

players. Subsequent studies are needed to develop the

system's full potential, broaden the investigation's

sampling, and examine all game fundamentals.

ACKNOWLEDGEMENTS

The authors are grateful to "Saroli Club" Castel

Gandolfo (Rome – Italy) for permitting them to take

the measurements within their facilities. The

availability and support provided in the research to

the Saroli family, coach Matteo Petrolati, and the

participants of the study

REFERENCES

Alcala, E., Voerman, J., Konrath, J., & Vydhyanathan, A.

(2021). Xsens DOT wearable sensor platform white

paper. White Paper.

Annino, G., Bonaiuto, V., Campoli, F., Caprioli, L., Edriss,

S., Padua, E., Panichi, E., Romagnoli, C., Romagnoli,

N., & Zanela, A. (2023). Assessing Sports

Performances Using an Artificial Intelligence-Driven

System. 2023 IEEE International Workshop on Sport,

Technology and Research (STAR), 98–103.

Bertino, L., Mohovich, B., & Pankhurst, A. (2012).

Manuale Performance Workshop 2012. Professional

Tennis Registry.

Caprioli, L., Campoli, F., Edriss, S., Frontuto, C., Najlaoui,

A., Padua, E., Romagnoli, C., Annino, G., & Bonaiuto,

V. (2024). Assessment of Tennis Timing Using an

Acoustic Detection System. 2024 IEEE International

Workshop on Sport, Technology and Research (STAR),

285–289.

Casale, L. (2003). PHYSICAL TRAINING FOR TENNIS

PLAYERS. Professional Tennis Registry.

Castellani, A., D’Aprile, A., & Tamorri, S. (2007). Tennis

training. Società stampa sportiva.

Charmant, J. & contributors. (2021). Kinovea (0.9.5).

https://www.kinovea.org

Cohen, J. (2013). Statistical power analysis for the

behavioral sciences. routledge.

Cooper, H., Hedges, L. V., & Valentine, J. C. (2019). The

handbook of research synthesis and meta-analysis.

Russell Sage Foundation.

Delgado-García, G., Vanrenterghem, J., Ruiz-Malagón, E.

J., Molina-García, P., Courel-Ibáñez, J., & Soto-

Hermoso, V. M. (2021). IMU gyroscopes are a valid

alternative to 3D optical motion capture system for

angular kinematics analysis in tennis. Proceedings of

the Institution of Mechanical Engineers, Part P:

Journal of Sports Engineering and Technology, 235(1),

3–12.

Ebner, C. J., & Findling, R. D. (2019). Tennis stroke

classification: Comparing wrist and racket as imu

sensor position. Proceedings of the 17th International

Conference on Advances in Mobile Computing &

Multimedia, 74–83.

Edriss, S., Romagnoli, C., Caprioli, L., Zanela, A., Panichi,

E., Campoli, F., Padua, E., Annino, G., & Bonaiuto, V.

(2024). The Role of Emergent Technologies in the

Dynamic and Kinematic Assessment of Human

Movement in Sport and Clinical Applications. Applied

Sciences, 14(3), 1012.

Elliott, B. (2006). Biomechanics and tennis. British Journal

of Sports Medicine, 40(5), 392–396.

Elliott, B., Takahashi, K., & Noffal, G. (1997). The

influence of grip position on upper limb contributions

to racket head velocity in a tennis forehand. Journal of

Applied Biomechanics, 13(2), 182–196.

Fleisig, G., Nicholls, R., Elliott, B., & Escamilla, R. (2003).

Tennis: Kinematics used by world class tennis players

to produce high-velocity serves. Sports Biomechanics,

2(1), 51–64.

Fu, M. C., Ellenbecker, T. S., Renstrom, P. A., Windler, G.

S., & Dines, D. M. (2018). Epidemiology of injuries in

tennis players. Current Reviews in Musculoskeletal

Medicine, 11, 1–5.

Hernández-Belmonte, A., & Sánchez-Pay, A. (2021).

Concurrent validity, inter-unit reliability and biological

variability of a low-cost pocket radar for ball velocity

measurement in soccer and tennis. Journal of Sports

Sciences, 39(12), 1312–1319.

Impact Distance Detection in Tennis Forehand by an Inertial System

281

Johnson, C. D., & McHugh, M. (2006). Performance

demands of professional male tennis players. British

Journal of Sports Medicine, 40(8), 696–699.

Knudson, D., & Elliott, B. (2004). Biomechanics of tennis

strokes. In Biomedical engineering principles in sports

(pp. 153–181). Springer.

Kos, M., & Kramberger, I. (2018). Tennis stroke

consistency analysis using miniature wearable IMU.

2018 25th International Conference on Systems,

Signals and Image Processing (IWSSIP), 1–4.

Lambrich, J., & Muehlbauer, T. (2023). Biomechanical

analyses of different serve and groundstroke techniques

in tennis: A systematic scoping review. PLoS One,

18(8), e0290320.

Landlinger, J., Lindinger, S., Stöggl, T., Wagner, H., &

Müller, E. (2010). Key factors and timing patterns in

the tennis forehand of different skill levels. Journal of

Sports Science & Medicine, 9(4), 643.

Martin, C., Sorel, A., Touzard, P., Bideau, B., Gaborit, R.,

DeGroot, H., & Kulpa, R. (2021). Influence of the

forehand stance on knee biomechanics: Implications for

potential injury risks in tennis players. Journal of Sports

Sciences, 39(9), 992–1000.

Nesbit, S. M., Serrano, M., & Elzinga, M. (2008). The role

of knee positioning and range-of-motion on the closed-

stance forehand tennis swing. Journal of Sports Science

& Medicine, 7(1), 114.

Reid, M., & Elliott, B. (2002). Tennis: The one-and two-

handed backhands in tennis. Sports Biomechanics, 1(1),

47–68.

Reid, M., Elliott, B., & Crespo, M. (2013). Mechanics and

learning practices associated with the tennis forehand:

A review. Journal of Sports Science & Medicine, 12(2),

225.

Roetert, E. P., & Kovacs, M. (2019). Tennis anatomy.

Human Kinetics.

Seeley, M. K., Funk, M. D., Denning, W. M., & Hager, R.

L. (2016). Tennis forehand kinematics change as post-

impact ball speed is altered. In The Biomechanics of

Batting, Swinging, and Hitting (pp. 181–192).

Routledge.

Zanela, A., Caprioli, L., Frontuto, C., & Bonaiuto, V.

(2024). Integrative Analysis of Movement: AI-

Enhanced Video and Inertial Sensors in Athletic

Contexts. International Workshop on Engineering

Methodologies for Medicine and Sport, 625–641.

icSPORTS 2024 - 12th International Conference on Sport Sciences Research and Technology Support

282