SPECTRAL ANALYSIS OF A HUMAN WALKING SEQUENCE
USING MEDIAL STEREO IMAGES
Masako TSURUOKA, Ryosuke SHIBASAKI, Shunji MURAI
Institute of Industrial Science, University of Tokyo,
7-22-1, Roppongi, Minato-ku, 106 Tokyo, JAPAN
Tel: +81-3-3402-6231 Ext. 2560, Fax:+81-3-3479-2762
E-mail: masako(shunji.iis.u-tokyo.ac.jp
Masahiro KURITA
2nd Dept. of Internal Medicine, Faculty of Medicine, University of Tokyo, JAPAN
Takao WADA
Dept. of Internal Medicine, Inagi Municipal Hospital, JAPAN
Elgene O. BOX
Dept. of Geography, University of Georgia, USA
Commission V, Working Group 5
KEY WORDS: Biometrics, Spectral Analysis, AR-Modeling, 3-D Videogrammetry
ABSTRACT
The 3-D quantitative understanding of human movements suggests powerful new knowledge in medical
rehabilitation and physical fitness. Serial 3D measurement data of movement provide an important
information for new medical evaluation of each person's stability in daily movement. This study presents
a new approach based on spectral analysis and the utilizing 3-D videogrammetry of the walking
sequence of normal persons.
This study is based on three items. First is a synchronization system composed of two CCD video
cameras, recorders and a time-generator, which solve the stereo-image matching. It recorded
simultaneously a real-time control symbol on each right and left serial image. Second is a simple 3-D
videogrammetric system, named "New Bird Man”, for non-metric cameras using bundle adjustment with
self-calibration. Third is spectral and also biomechanical analysis for finding person's walking features.
Case À Case B Case C Stereo Imae
da Er OM 4
Pere «c ACER
Pho. 1 Normal leg Pho.2 Stressed rt.leg Pho. 3 Damaged lt. leg
(experimentally fixed by orthosis) (experienced a broken leg)
566
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B5. Vienna 1996
The
expect
rehabi
which
dynan
develc
biostei
calibra
of mi
obtain
Fig
Fig: 2
fixed
image
(Case
difficu
featur
fluctu:
relatio
new
utilizii
Model
2.
cus
