Comparison of Camera based and Inertial Measurement Unit based

Motion Analysis

Seongho Jang

, Si-bog Park

, Sang-bok Moon

Jae Min Kim

and Shi-uk Lee

Department of Rehabilitation Medicine, Hanyang University, Seoul, Korea

Department of Neurosurgery, Hanyang University, Seoul, Korea

Department of Rehabilitation Medicine, Seoul National University Boramae Hospital, Seoul, Korea

Keywords: Motion Analysis, Inertial Measurement Unit, Gait.

Abstract: Camera-based 3D motion analyzers are widely used to analyze body movements and gait, but they are

expensive and require a large dedicated space. This study investigated whether inertial measurement unit

(IMU)-based systems can replace such systems by analyzing kinematic measurement parameters. IMUs

were attached to the abdomen and thigh and the shank and foot of both legs. The participant completed a 10

m-gait course 10 times and the hips, knees, and ankle joints were observed from the sagittal, frontal, and

transverse planes during each gait cycle. The experiments were conducted with both a camera-based system

and an IMU-based system. The measured gait analysis data were evaluated for validity and reliability using

RMSE. In this regard, the differences between the RMSE values of the two systems determined through

kinematic parameters ranged from a minimum of 1.39 to a maximum of 3.86. These results confirm that

IMU-based systems can reliably replace camera-based systems for clinical body motion analysis and gait

analysis.

1 INTRODUCTION

The level of improvement in gait and quantification

of body motion corroborate clinical decisions in the

treatment process and is used for functional

assessment in clinical gait analysis and

rehabilitation. Interest in gait evaluation and gait

improvement is increasing for non-patients as well

as young persons who have abnormal gait. Gait

analysis has evolved from a simple 2D video camera

analysis to optical motion capturing using several

infrared cameras and 3D motion analysis systems.

The 3D motion analyzers currently widely used for

gait analysis record body motion by reading location

coordinate values of body markers attached to in

body in real time with several infrared cameras in a

limited space.

However, both the purchase price and

maintenance of these motion analyzers are high.

Further, in order to take measurements from various

angles, several cameras and much space are

required. In addition, because such systems have to

be installed by professionals and require complex

setup and preparation for experiments and data

analysis, they are difficult to apply in clinical

settings.

Under different experimental conditions and

environments, the measurements obtained can also

differ based on the setting’s characteristics.

Consequently, issues concerning the validity and

reliability of the measurements obtained from these

machines also exist. With the aim of developing

systems that are without the disadvantages outlined

above, in recent times, research has been focused on

gait analysis using inertial measurement units

(IMU).

Recent advancements in sensor technology

enable simple and economic analyses to be

performed using IMUs. The inertial sensors usually

comprise a gyroscope, an accelerometer, and a

magnetometer, which enable economical

measurements of gravitational force and

acceleration. Changes in the Euler angle, yaw, pitch,

and angles of rolling axis can also be measured

using the gyroscope.

Numerous studies on gait analysis using inertial

sensors have focused on detection of gait phase, and

measurement of joint angles, segment angles, and

stride lengths. The results of these studies indicate

Jang, S., Park, S., Moon, S., Kim, J. and Lee, S.

Comparison of Camera based and Inertial Measurement Unit based Motion Analysis.

DOI: 10.5220/0006716601610167

In Proceedings of the 7th International Conference on Sensor Networks (SENSORNETS 2018), pages 161-167

ISBN: 978-989-758-284-4

161

that wireless inertial sensor systems on the lower

body could analyze and evaluate the characteristics

of gait.

However, gain analysis data from the use of

inertial sensor systems are scarce. Furthermore,

because the technology is not considered fully

complete, inertial sensor systems are not widely

used for clinical gait analysis. Furthermore, their

accuracy is in doubt. In this study, the accuracy of

IMU-based sensor systems was investigated through

spatial-temporal and kinematic parameters on the

same subject and comparison with results from

camera-based 3D motion capture systems to

determine whether IMU-based systems can replace

camera-based systems.

In this study, a gait analysis system that analyzes

and quantifies the kinematic data of a specific part of

body was developed. Further, measurements

obtained from wearable IMU sensors on the lower

limb were compared to those from a camera-based

optical motion capture (OMC) system and their

validity evaluated. In addition, tests were conducted

in multiple settings to confirm the reliability and

effectiveness of IMUs. To the best of our

knowledge, no studies have reported on the

reliability of IMUs. Thus, confirmation of IMU

effectiveness and its accuracy will provide important

reference data for further studies in related fields.

The developed system can be applied in clinical and

rehabilitation settings.

2 MATERIALS AND METHOD

2.1 Participant and Gait Measurement

The subject of this study was a healthy adult male

with no musculoskeletal disabilities (age: 40s,

height: 180 cm, weight: 90 kg). The experiment was

conducted in three different hospitals (National

Rehabilitation Center, Veterans Health Service

Medical Center and Yonsei University Hospital)

between March 2016 and May 2016.

Each hospital had all the necessary equipment to

simultaneously conduct gait pattern analysis for both

the camera-based and IMU-based systems (Fig. 1).

In addition, the procedures in this test were

performed with the approval of Hanyang University

Guri Hospital (IRB File No. GURI 2015-03-001-

003).

The participant completed a 10 m-gait course 10

times in each experimental setting. During the gait

cycle, the kinematic parameters of hip, knee, and

ankle joint were inspected from the sagittal, frontal,

and transverse planes. In addition, the temporal-

spatial parameter was inspected. All experimental

trials were conducted in identical conditions.

Figure 1: Camera-based system and IMU-based system.

2.2 Experiment Equipment and

Procedure

For gait pattern analysis, the camera-based system

comprised VICON MX-T10 (Vicon Motion Systems

Ltd., Oxford, UK), which is the most widely used

system, and Motion Analysis (Raptor-E Digital Real

Time System; Motion Analysis, Santa Rosa, CA,

USA). The IMU (35 mm × 60 mm × 25 mm)-based

gait analysis system (Motion Track, R. Biotech Co.,

Ltd., Seoul, Korea) consisted of gyroscope,

accelerometer, and magnetometer sensors.

To evaluate the validity of the IMU, a reflective

marker-based 3D infrared camera system was

simultaneously used. The markers, used to analyze

the lower limb motion during gait were attached to

body using the plug-in-gait marker set method.

Wearable wireless IMUs were attached to the

abdomen, femur, tibia, and foot of both legs and

affixed with stretch bands. As shown in Fig. 2, the

IMU sensors were placed on a holder to increase

stability and accuracy (Fig. 2).

Each sensor’s signal was received and collected

using Bluetooth communication. The spatio-

temporal (gait cycle time, stance, swing phase,

velocity, distance, etc.) and kinematic (hip, knee,

ankle angle in three dimensions) data were

calculated using MATLAB

(ver. 2010a,

MathWorks Inc., USA). Before the actual

measurements, the participant underwent several

SENSORNETS 2018 - 7th International Conference on Sensor Networks

162

trials with the markers and IMUs attached in order to

familiarize himself with the gait conditions.

The validity of the gait analysis was analyzed

using the root mean square error (RMSE) of

parameters simultaneously obtained through the

camera-based system and the IMU-based system.

The reliability of the IMU was inspected using the

RMSE of kinematic parameters measured with the

IMU in three different experimental settings with a

certain time interval between each.

Figure 2: Hardware design of the IMU (the sensor was

placed in the holder).

2.3 Attitude and Heading Reference

Systems (AHRS) Module

In this study, inertial sensor-based AHRS was

designed and developed. When attached to the body

joints, the AHRS could measure the kinematic

motions of each joint objectively. In addition, the

AHRS measured the direction of the gravity and

magnetic field of the earth.

The AHRS module is composed of inertial

sensor, a microcontroller for receiving and

processing the signals, a Bluetooth module for

communication, and a battery charging circuit.

The inertial sensor used for the module in this

study was an integrated sensor (MPU9250,

Invensense, USA) composed of a gyroscope (range

± 2000 °/s), an accelerometer (range ± 16 g), and a

magnetometer (range ± 49 G). The signals were

programmed to be transmitted to the microcontroller

through SPI communication at 100 Hz in each

signal. The collected angular velocity, acceleration,

and magnetometer values were combined and the

gradient descent algorithm used to calculate the

Euler angle, yaw, pitch, and roll of the AHRS

module. The calculated values were transmitted to a

PC using a wireless Bluetooth module (PAN1321i,

Panasonic, Japan).

On the basis of the data from the magnetometer,

which provides data on the earth’s magnetic field

using the gradient descent algorithm, and data from

the accelerometer, which provides data on the

gravity and inertia, the gyroscope’s inaccurate

measurement of angular velocity was supplemented

and integrals were conducted to calculate and

reliably determine the Euler angle.

2.4 Gait Event Detection and Temporal

Parameters Calculation

The differential calculated from the foot’s Euler

angle determined the gait event, as shown in Fig. 3,

and the temporal parameters of gait were also

suggested. Fig. 3 explains the algorithm used to

determine the temporal parameter using gyroscope

on the foot. The figure shows the quantification of

the gyroscope features during the gait cycle of each

foot observed from the sagittal plane.

Figure 3: Graphical illustration of the algorithm used to

determine the gait event using rotation angle on the foot.

The inertial sensor data based on the verified

algorithm detected the heel strike (HS) and toe-off

(TO) points.

Peak rotation rate is the maximum achieved

rotation rate of the ankle during the swing phase.

The minimum value of TO is the minimum value

larger than the peak rotation value at mid-swing. In

addition, at HS, the peak of the negative rotation

value is observed at the first minimum after the

maximum rotation rate during the mid-swing period.

After the detection of HS and TO, the gait cycle

was formed to calculate gait temporal parameters.

Based on these time events, the temporal parameters

swing phase and stance phase can be calculated

using Eqs. (1) and (2), respectively.

Comparison of Camera based and Inertial Measurement Unit based Motion Analysis

163

Swing Phase, SW:













 



  (1)

Stance Phase, :

















  (2)

2.5 Spatial Parameters Calculation

In spatial parameters calculation, the distance

traveled by the subject is determined by the double

integration of momentary accelerations

measurements. The position or distance values

obtained by the integration is known to be suitable

only for a short term because of the drift error of the

accelerometer. In other words, calculation of

velocity and distance using the double integration of

acceleration measurements produced a relatively

large accumulated error. To avoid this accumulated

error, the values should be measured after every step

the pedestrian takes. If the velocity and distance

estimations are measured in each step, the

successive measurements of speed and distance are

not affected. Therefore, the accumulated error of the

AHRS was corrected. Fig. 4 shows a schematic

diagram of the two-phase cumulative error reduction

algorithm used to minimize the accumulated error of

the double integration, which was calculated using

the acceleration values and angular velocity from the

AHRS modules.

Velocity and distance were calculated by double

integrating the acceleration measurements. The

gravitation influences and accumulated error were

removed in order to calculate accurate values.

Figure 4: Spatial parameters calculation algorithm.

2.6 Calculation of Joint Angles

In order to measure the joint angles during

rehabilitation gait analysis, a total of seven AHRS

system modules were attached to the participant’s

joints. The modules were attached to the abdomen,

bilateral femurs, bilateral tibias, and the feet using a

stretch band, and the angle joints were calculated

using the Euler angles obtained from each joint. An

algorithm to calculate joint angles, which are

important biological measurement for rehabilitation,

was also developed in this study.

Fig. 5 shows a conceptual map of the algorithm

that calculates the joint angle in each segment. The

conceptual map uses the joint angle between the

femur and the tibia as an example. The example

shows the method used to calculate the angle

between the femur and tibia; the same algorithm can

be applied to other segmental joint angles.

Figure 5: Conceptual map of segmental joint angle

calculation method.

Fig. 5 represents tibia anatomical (TA) and femur

anatomical (FE), and each sensor axis was labeled as

tibia measurement (TM) and femur measurement

(FM). To convert the axis of each sensor into one

single axis, conversion matrix T

TATM,

which is a

matrix in which the sensor axis is converted to tibial

axis, and T

TAFM,

which is a matrix in which the

sensor axis is converted to femoral axis, were used.

The ultimate matrix that representing the joint angle

between the two sensors is expressed in Eq. (3).



































  

  

  

 (3)

In Eq. (3), S represents the alignment matrix

between two axis sensors of the earth (FE and TE)

while M

and M

represent the direction of the femur

and tibia relative to the earth axis. As shown in Eq.

(4), T

FATA

matrix terms were used to calculate the

joint angles flexion/extension, abduction/adduction,

and internal/external rotations:





























)









 (4)

SENSORNETS 2018 - 7th International Conference on Sensor Networks

164





























3 RESULTS

In order to evaluate the performance of the gait

analysis system, the segmental joint angles on both

lower limbs were measured with seven AHRS

system modules. The attachment locations were the

abdomen, bilateral femurs and tibias, and the feet.

Based on the Euler angles of each joint provided by

the individual modules, the joint angles were

calculated with the joint angle calculation algorithm.

The Euler angles obtained from the AHRS in

segmental joints while the participant completed the

10 m-gait course were used to calculate joint angles

and 10 trials were conducted for measurement under

identical protocol.

Table 1 shows the inspection results for the

validity of the IMU-based system. The validity was

evaluated by comparing the temporal and spatial

parameters of the gait measured by the camera-based

and IMU-based systems in the three separate

hospitals.

The velocity measured with the IMUs was in the

range 1.16–1.20 m/s, whereas that measured with

the camera-based system was in the range 1.23–1.31

m/s. The stride lengths measured with the IMUs and

the camera-based system were in the ranges 1.15–

1.27 m and 1.19–1.32 m, respectively.

The stance phase (%) measured with the IMUs

was in the range 56–58%, whereas they were in the

range 61–63% for the camera-based system. The

swing phases (%) measured with the IMUs and the

camera-based system were in the ranges 42–44%

and 37–39%, respectively. Overall, the values

measured with the two systems did not show

significant differences.

Table 2 shows the inspection results for the

validity related to kinematic parameters of the IMU-

based system obtained by comparing the gait data

measured with camera-based and IMU-based

systems in three different hospitals. In order to

analyze the accuracy of the IMUs, the differences in

lower limb joint angles measured with the two

systems are shown. On the basis of the Euler angles

obtained from the AHRS modules on each body

segment, the segmental joint angles during gait cycle

were calculated. The values were processed in a 3D

space on the sagittal, frontal, and transverse planes.

Table 1: Temporal and spatial parameters of the camera-

based system and the IMU-based system obtained in the

three separate hospitals.

Velocity

(m/s)

Stride

length

(m)

Stance

phase

(%)

Swing

phase

(%)

National

Rehabilitatio

n Center

IMU

1.20

1.23

Vicon

1.31

1.19

Veterans

Medical

Center

IMU

1.18

1.27

Motion

analysis

1.25

1.32

Yonsei

University

Hospital

IMU

1.16

1.15

Vicon

1.23

The RMSE values of each of the sagittal, frontal,

and transverse planes are shown in Table 2. The

average RMSE value of the ankle joint angle on the

frontal plane was the lowest at 1.60. In addition, the

RMSE value of the ankle joint angle on the

transverse plane was the highest at 3.82. The results

verify the validity of the IMUs. The RMSE values

obtained at the hospitals are shown in Table 2.

Table 2: Kinematic parameters for the camera-based

system and the IMU-based system obtained at the three

hospitals.

National

Rehabilitat

ion Center

Veterans

Medical

Center

Yonsei

University

Hospital

Average

RMSE

Sagittal

1.80

1.46

2.14

1.80

2.87

2.37

2.91

2.72

1.63

2.26

1.43

1.77

Frontal

3.48

1.57

2.68

2.58

1.77

2.13

2.08

1.99

1.39

1.81

1.60

Transverse

3.86

3.72

4.25

3.94

2.99

3.01

2.95

2.88

3.35

5.23

3.82

H: RMSE of Hip Joint angle, K: RMSE of Knee Joint

angle A: RMSE of Ankle Joint angle

4 DISCUSSION

It has been reported that the mechanical accuracy of

IMUs could produce errors when measuring body

movements such as joint angle measurements.

During the measurement of acceleration and angular

velocity, the measurement plane of the IMU

Comparison of Camera based and Inertial Measurement Unit based Motion Analysis

165

modules on the body did not mechanically coincide

because of the curves on the body.

In this study, a camera-based system was

simultaneously used with the IMU-based system.

The differences between the RMSE values of the

two systems determined through kinematic

parameters with a tolerance close to 1%. Therefore,

the comparison results of two systems indicate that

IMU-based systems can replace camera-based

systems. The errors in joint angles during gait

analysis are within the tolerance range and the errors

could be reduced by replacing the gyroscope,

accelerometer, and magnetometer sensors with an

integrated sensor.

Further, the RMSE values of the kinematic

parameters measured with the IMU-based systems in

the three different experimental settings with a

tolerance close to 1%. Therefore, it can be inferred

that IMU-based systems are reliable for gait

analysis. Compared to the 2% error rate reported by

previous studies that used relatively more expensive

sensors, this study showed similar performance with

those studies that used high-cost sensors.

The limitations of this study include the fact that the

study was conducted on one participant and the

measurement session was extended over a long

period. Although the healthy participant tried to

maintain his health and physical activities for three

months during the experimental trials, the

measurements in different hospitals were taken over

an extended period.

Further studies on IMU-based gait analysis will

attract increased attention and demand. Therefore, a

system that provides feedback for gait correction and

evaluation will be developed in future work..

5 CONCLUSIONS

Gait analysis is currently conducted very rarely

owing to high equipment cost, complex procedure,

and space restriction. Therefore, an IMU-based

system was inspected to verify its validity and its

potential to replace camera-based systems. The

results indicate that IMU-based systems can be

effectively used in clinical settings and could be

applied to other fields that require gait analysis.

Furthermore, it is expected to be widely distributed

in related fields. Because IMU-based systems

provide accurate gait data in real time, they could

contribute to faster diagnosis and evaluation by

physicians.

This study verified the validity and the reliability of

IMU-based systems. The results indicate that IMU-

based systems can be widely used for rehabilitation

and gait analysis in clinical settings. It will be

necessary to develop interaction-coaching systems to

improve the accessibility of such systems. In

addition, a new type of gait analysis system that

portrays gait data as graphs, 3D avatars, and

webcams should be developed. The development of

IMU-based systems is expected to improve the

quality of patients’ lives as the cost for gait analysis

will consequently decrease.

ACKNOWLEDGEMENTS

This work was supported by Institute for

Information & communications Technology

Promotion(IITP) grant funded by the Korea

government(MSIP) (2017-0-01800).

REFERENCES

Simon, S. R., 2004. “Quantification of human motion:

Gait analysis - benefits and limitations to its

application to clinical problems,” Journal of

Biomechanics, Vol. 37, No. 12, pp. 1869-1880,

Sutherland, D. H., 2002."The evolution of clinical gait

analysis: Part II kinematics," Gait & Posture, Vol. 16,

No. 2, pp. 159-179,

Sutherland, D. H., 2005. "The evolution of clinical gait

analysis part III – kinetics and energy assessment,"

Gait & Posture, Vol. 21, No. 4, pp. 447-461,

Ugbolue, U. C. et al., 2013."The evaluation of an

inexpensive, 2D, video based gait assessment system

for clinical use," Gait & Posture, Vol. 38, No. 3, pp.

483-489,

Costigan, P. A. et al., 2002."Knee and hip kinetics during

normal stair climbing," Gait & Posture, Vol. 16, No.

1, pp. 31-37,

Favre, J. et al., 2008."Ambulatory measurement of 3D

knee joint angle," Journal of Biomechanics, Vol. 41,

No. 5, pp. 1029-1035,

Watanabe, T. et al., 2011."A preliminary test of

measurement of joint angles and stride length with

wireless inertial sensors for wearable gait evaluation

system," Computational Intelligence and

Neuroscience, Vol. 2011, Article No. 6,

Zheng, H. et al., 2005."Position-sensing technologies for

movement analysis in stroke rehabilitation," Medical

and Biological Engineering and Computing, Vol. 43,

No. 4, pp. 413-420,

Cloete, T., and Scheffer, C., 2008 "Benchmarking of a

full-body inertial motion capture system for clinical

gait analysis," in 2008 30th Annual International

Conference of the IEEE Engineering in Medicine and

Biology Society, pp. 4579-4582, IEEE,.

SENSORNETS 2018 - 7th International Conference on Sensor Networks

166

Boudarham, J. et al., 2013. "Variations in kinematics

during clinical gait analysis in stroke patients," PloS

one, Vol. 8, No. 6, pp. e66421,

Karasawa, Y. et al., 2013. "A trial of making reference

gait data for simple gait evaluation system with

wireless inertial sensors", in 2013 35th Annual

International Conference of the IEEE Engineering in

Medicine and Biology Society, pp. 3427-3430, IEEE,

Van Acht, V. et al., 2007. "Miniature wireless inertial

sensor for measuring human motions," in 2007 29th

Annual International Conference of the IEEE

Engineering in Medicine and Biology Society, pp.

6278-6281, IEEE,

Luinge, H. J., and Veltink, P. H., 2005. "Measuring

orientation of human body segments using miniature

gyroscopes and accelerometers," Medical and

Biological Engineering and Computing, Vol. 43, No.

2, pp. 273-282,

Zhu, R., and Zhou, Z., 2004. "A real-time articulated

human motion tracking using tri-axis inertial/magnetic

sensors package," IEEE Transactions on Neural

Systems and Rehabilitation Engineering, Vol. 12, No.

2, pp. 295-302,

Kim, M.-S. et al., 2014. "Development of a high-precision

calibration method for inertial measurement unit,"

International Journal of Precision Engineering and

Manufacturing, Vol. 15, No. 3, pp. 567-575,

Comparison of Camera based and Inertial Measurement Unit based Motion Analysis

167