System for tracking markers applied for kinematical human motion analysis
Abstract
In this talk, I'm going to present a system for automatic tracking of
markers, that we developed in the Laboratory of Instrumentamion for
Biomechanics, University of Campinas, Brazil.
This system is based on image sequence captured by a normal CCD camera and
image processing algorithms. Markers can be defined as objets that are
attached or fixed to human body. These markers help to track the movement
of interested points of the body, for example, the joints.
The importance of such systems is related to their application in
different areas, such as clinical gait analysis, orthopedic, neurology,
sports, etc.
In particular, commercial optoelectronic system based on markers
and special cameras have been extensively used to record 3-D kinematics of
the human motion. The tracking of the markers in most of these systems is
accomplished using special purpose dedicated hardware and the applications
are limited to a controlled environment.
In our system we used the i.age-processtng tools, in particular, the
mathematical morphology to treat the problem of tracking in an
uncontrolled environment.
I'm going to show the sequence of algorithms that we developed and tested
for different markers and different type of movements, such as walking,
running, jumping, etc.