The Application of Motion Pattern Recognition in the Behaviometrics of Human Kinematics

Andar Bagus Sriwarno

Abstract

Current trends seem to accredit motion as a sensible biometric feature for human identification in spite of
physiological parameter which related to the shape of the body. As in many image processing system, the
recognition is running under the circumstances of the lighting condition that may moving object acquitioned by
the computer. Through this principes, this paper presents the application of video inter-frame analysis system
for determining human motion from the sagittal view video sequences. By using this method, the behavior of
person can be recognized and determined the type of motion. Some researches have coined the term
behaviometrics for this class of biometrics. The experiment was conducted by asking subjects to conduct the
simple motion. Eight healthy male subjects participated in the experiment to raise from the deep squatting (STS)
at the three different foot elevations: full squatting (FS) which the plantar feet were fully contact to the floor (0
degree), slope squatting (SS) which a 15 degrees of wedge was applied underneath the feet, and tiptoe squatting
(TS) which subjects rose the heel by extending the metatarsal joint at their preferred height. The result showed
that raising the heel significantly affected the postural adjustment that at TS condition was significantly smaller
in forward displacement on the knee and the hip during ankle dorsiflexion and influenced to short the time taken
at the forward movement to conduct STS movement than shown at FS and SS. Significant differences were also
appeared in the parameter of task period and segmental body angle. These findings suggest that the differences
way to achieve STS movement in some ways can contribute to decrease task period and to differ the pattern of
motion that furthermore supports the identification in behaviometrics.

Keywords: behaviometrics, motion pattern, video acquitition.

Full Text:

PDF