Implementation of a Real-Time Human Movement

Classifier Using a Triaxial Accelerometer for

Ambulatory Monitoring

Presenter: Wei-Chen Lin Adviser: Dr. Cheng-Jui Hung

2009/2/25

2

Outline

• Introduction• Paper review• Motivation and Purpose

• Materials and Methods

• Results

• Future works

• References

3

Introduction

Triaxial accelerometer applications

• Numerical Analysis

• Experiment

• Behavior

• Monitoring systems

4

Paper review(1)

Purpose• Advances in miniature sensor and wireless technologies

have resulted in interest in the development of systems for monitoring subjects over long periods of time using wearable monitoring units.

• From: Classification of basic daily movements using a triaxial accelerometer Author(s): Mathie MJ, Celler BG, Lovell NH, et al.Source: Medical & Biological Engineering & Computing   Volume: 42   Issue: 5   Pages:

679-687   Published: SEP 2004

5

Paper review(1) materials and methods

• Waist-mounted triaxial accelerometer– The unit was composed of two orthogonally mounted

biaxial accelerometers*(range± 10 g; frequency response: 0-500 Hz)

• Presented a more systematic approach to classification, based on a formal, hierarchical, decision tree.

• The algorithms were developed and tested using data collected from 26 normal, healthy subjects (seven female, 19 male; mean age 30.5 years-4-6.3 years standard deviation)

6

Paper review(1) flow chart

TA signal

Fall?

Transition?

Activity?

Upright-Upright?

Upright-Lying?

Lying-lying?

Lying-Upright?

Sitting?

Upright?

Lying?

Walking?

Lying faceDown?

Lying on left side?

lying on back?

Fall:raisealarm

walkingOther

movement

sittingUpright-to-

Lying transitionLying-to-lying

transitionLying-to-upright

transitionLying face

downLying on

back

Inverted:Raise alarm

Lying on leftside

standingLying on right

side

No

No

No

No No

NoNo

No

No

No

No

No

No

Yes

Yes

YesYes

YesYes

Yes Yes YesYes

No

Yes

YesYesYes

Yes

TA: Triaxial Accelerometer 三軸加速度器

Activity Rest

Level 1

Level 2

Level 3

Level 4

7

Paper review(1) results

Fig.1 Experimental Statistics tables

8

Paper review(1) conclusion

• Using this framework, a classifier for the identification of basic movements, based on a monitoring system consisting of a 686 single, waist-mounted triaxial accelerometer, was developed, in laboratory studies in which 26 subjects performed a specific routine of movements, the system obtained an overall sensitivity of 97.7% and specificity of 98.7% over a data set of 1309 movements.

9

Paper review(2)

Purpose• Step lengthcan be computed by means of a biaxial

accelerometer and a gyroscope on the sagital plane.

• From: Alvarez, J.C.; Gonzalez, R.C.; Alvarez, D.; Lopez, A.M.; Rodriguez-Uria, J.;Engineering in Medicine and Biology Society, 2007. EMBS 2007. 29th Annual International Conference of the IEEE

22-26 Aug. 2007 Page(s):5719 - 5722

10

Paper review(2) materials and methods

• Biaxial accelerometer and a gyroscope

• Motion is computed, at every stride, by estimating the distance traveled by the foot that swings forward on the air.

g-cosa-senaA

sena-cosaA

zxz

zxx

Fig.2 Experimental device

11

Paper review(2) experiments results

Fig. 3-4 One foot displacement, signals from the gyroscope (up), itsintegration (middle) and the corrected accelerations (down). Computations

are made with equations (2) and (3).

ω:角加速度 θ:角度

12

Paper review(2) experiments results

Fig. 5 Integrating the gyroscope signal ofthe sagital plane

13

Paper review(2) conclusion

• We have presented a method to estimate the step lengthbased on inertial feet attached sensors. Contrary to similarworks, a multisensor approach is applied in order to reduceuncertainty and to produce better estimations. An adaptedkalman filter based sensor fusion system is proposed. Initialresults are encouraging. Ongoing extended field experimentshave been designed to validate and generalize the results fora heterogeneous populations and walking conditions.

14

Motivation and Purpose

• Monitoring of human movement• Measured distance

– Design platform for measuring the distance

• To detect the occurrence of falls

15

Materials and Methods

• ST LIS302DL– 3-Axis– range : ± 8g– frequency response:100Hz or 400Hz

• Measurement platform– 28cm × 21cm × 3.7cm

• Microchip APP009– Microchip Dsp30F4011

16

Measurement platform

尺規刻度

可移動方向

APP009 實驗版

三軸加速度器

17

Results

-20

-10

0

10

20

30

40

1 45 89 133 177221 265309353 397441485 529573 617661705 749

Time(10ms)

(cm

/s2)

加速

度 1數列

Fig.7 速度曲線圖

-40

-30

-20

-10

0

10

20

30

1 45 89 133 177221 265309353 397441485 529573 617661705 749

Time(10ms)

(cm

/s)

速度 1數列

Fig.6 加速度曲線圖

18

Results

Fig. 8距離曲線圖

-7

-6

-5

-4

-3

-2

-1

0

1

2

3

1 44 87 130173216259302345388 431474517560603646689732

Time(10ms)

(cm

)距

離 1數列

19

Results

• The experimental results show that estimated value is measured 9 cm, the actual measurement of 7.8 centimeters. Error value of 11%

Ideal distance

number of measure

experimental

average

accuracy

9 cm 54 7.8 cm 89%

20

Future works

1. Paper review

2. Increase the distance measurement accuracy

3. Data collection and statistics

4. Reduce the board 5. Functional integration

21

References

• [1] C.V. Bouten,K. T.Koekkoek, M.Verduin, R.Kodde, and J. D. Janssen, "A triaxial accelerometer and portable data processing unit for the assessment of daily physical activity," IEEE Trans. Biomed. Eng., vol. 44, no. 3, pp. 136–147, 1997.

• [2] M. J. Mathie, A. C. F. Coster, N. H. Lovell, and B. G. Celler, "A pilot study of long term monitoring of human movements in the home using accelerometry," IEEE Trans. Biomed. Eng., vol. 10, pp. 144–151, 2004.

• [3] M. Makikawa, D. Murakami, " Development of an ambulatory physical activity and behavior map monitoring system," in 18th Annual Conf. IEEE Engineering in Medicine Biology Soc. Amsterdam, Holland, 1996.

• [4] M. J.Mathie, N. H. Lovell, A. C. F. Coster, and B. G. Celler, "Determining activity using a triaxial accelerometer," in Proc. 2nd Joint EMBS-BMES Conf., Houston, TX, 2002.

• [5] M. J.Mathie, B. G. Celler, N.H. Lovell, et al. "Classification of basic daily movements using a triaxial accelerometer," Medicine & Biological Engineering & Computing., vol. 42  pp. 679-687, 2004.

• [6] W. Zijlstra and A. Hof, "Assessment of spatio-temporal gait parameters from trunk accelerations during human walking," Gait & Posture., vol. 18, pp. 1-10, 2003.