Page 67 - 2024F

60 UEC Int’l Mini-Conference No.53

ctivit
E
tion of l
y
ower limb muscle a
stima
y people using alking in elderl during w
ss mo
markerle
og
tion cap
ture technol
y
a,
e Theodor angpusri, Re Masaaki Hir oi, K w
ou
Oga
Hidetaka Okada
Introduction
M arkerle ss M o G ait A sis y nal
ture for
tion Cap
educ
all risk an
d r
es
lin
y Dec
e Mobilit
ases f
e in Elderly: In
cr
y of lif
.
e qualit
e s
aptur
e a mark
aluat
Objectiv
m e: Ev e yst erless m
otion c
) c
ed with An
nP
ombin
yBody musculo o se Ope (
sk
eletal
w
er limb musc e e lo odeling t m
ompar
o estimat
le activit
y an
d c
act eristics.

male gait char s. f e male v
e
er-based s
yst
an be ms c aditional mark r ds: T Gr oun
nsiv
erless appr e; a mark y oach ma som
xpe
e
e an
d cumber
on-inv
e , ac ernativ essible alt c asiv e er a n off
.
M ethods
P a rtic i p a nt ?
Figur e 4 . A g e v s. S c e tan P hase P c er nta e g e
or
ec
ects r
S ed elderly sub elf-select j alkingÜ ded during w
S ub j ects : 13 n ee , ages betw 7 7 - 91 (6 F , e male 7 M ).

ale
M a r k erless Motion C a p tur
o
se Ope nP : D ects et j ntify f oot om video t sitions fr oint po
o ide
ntsÜ oe-off e v e c ontact an d t
yBody
act
ed kin
matic data t
An : U le e musc o estimat xtr ses e
e
d or c es, an d r oun action f e or f es, gr c j oint loads.
oc
D a t a P essin (
r
c
ormali
Int erpolation : le n Each gait c y z om ed fr 0 – 100% an d
o esampled t r 101 data pointsÜ
le
d musc
e
Analysis : C wing tim
, s
ompar
e tim
c
, an
e
ed stan
ation pa activ tt s. ee erns betw
n ge
n
der
tion Cap
cle & Step C y M ture o P ss roce
e Figur 5. c n e in stan c e phase per c ntage b y se x e Diff e er
t Foo Cont (Ron) ct a Right ct (L on) t Cont a Left Foo
y ts cle Repea G C ait
DISCUSSION, C
ONCLUSION & FUTURE WORK
1 3 4 5 N on -I nv a e siv A n a lysis: C se an yBody d An o nP ombining Ope
2
er
erless m
o
st-eff
ectiv
, mark
er
w
o assess lo s a c off
ethod t
e
ction.

limb fun
t Off (L off) Left Foo Off (Roff) t Foo Right
c
ertain
es in c
onger stan
Figur e . 1 oc ess St e le & Motion Captur c p Cy e Pr or Elderly ations f Implic : L e tim
y in
at
dic
alking, while y-f dividuals ma e stabilit in
ocused w
d at egies an ation data c an inf ehabilitation str orm r le activ musc
ts
Re
sul
e all pr f ntion.

v e
F utur e R ese a r c h
L ar n male pe o dee ger samples t – male c ompar f e e ativ
insightsÜ
omat
M achin e L ormal e abn egr arning int e
ation t
o further aut
ection .
gait det
Figur e 2. M ar k er l ess M ot i on C apt u r e s
Reference
ptual Computing Lab
e
er
–Ü . er c sit y - P egie Mellon Univ Carn
/ s: / . A po se GitHub Re o y or sit nP
Ope
ailable at: http
v
npo ptual-Computing-Lab/ e github
ope
s´
om/CMU-P
er
c
.c
vÜ o Ope nP
e Multi-P
er
son 2D P
altim
se: Re
o
se Estimation Using
y Fields. IEEE T
init
P
ff
r
attern Analysis ansactions on P art A
eÉ d Machin
an
. A
e
g/ .or s/1812.0800i ab pÜ arxiv / / s: ellige Int
v
c
n
ailable at: http
m - Japan
yst
¤Ü e yBody Modeling S y ab err ation. An e Corpor t T
ab
y
err
.t
/
o t e /www v ailable at: http s: . A Distribut or
.jp/
.c
An
yBody/
Mail: e_79@hotmail.se
Theodor
e Figur male ) or ar old f e le ( c e 78 y y e during a gait c c 3.Musc le f