)F 
between 
er half of 
] with 10 
re 45°per 
S to the 
ned. 
ıterval of 
times are 
ames and 
onnected, 
  
e Monitor part 
+ Stereo image with times are displayed on VTR 
monitor. 
+ Time data, pan and tilt angles and tracking image 
are displayed on PC monitor. 
Then, the stereo image sequences, times, pan and tilt rotation 
angles can be obtained in real-time. 
3. GAIT EXPERIMENT 
In order to evaluate the road structure, the gait experiments for 
the young and the aged using HVT system, artificial sidewalk 
and the aged experience kit were performed. 
The zoom lenses (f=35mm) were fitted on the CCD1 and CCD2 
so that feature points on the lower of the body can be acquired. 
Table 1 shows the major components of the experiment. 
Table 1 Major components of the experiment 
  
  
  
  
  
Camera Sony XC-711 NTSC 
Lens 35mm(normal lens16mm) 
Height 7m 
Base line 0.8m 
  
  
  
3.1 The aged Experiment Kit 
The aged experience kit is developed so that anyone can 
experience the aged behaviour. Figure 2 shows the appearance 
of the aged experience kit, and the major features of the aged 
experience kit are as follows: 
A) It is possible to restrict movement of each joint. 
B) Furthermore, it is possible to keep the posture with the head 
bent forward. 
C) Attaching the goggles reproduces the aged field of view. 
  
Figure 2. Aged experiment kit 
3.2 Artificial Sidewalk 
8m artificial sidewalk was prepared for the gait experiments, 
and the sidewalk has ability to change the slope from 0% to 
10%. The gait analyses were performed for 2 slope modes 
(0%,5%). 
3.3 Auto Tracking Method 
Generally, dynamic analysis of human motion has been 
performed under the condition that camera position and rotation 
are fixed, and some markers are fitted on the feature points of 
the body such as knee, elbow, and so on. However, in order to 
analyze the most natural human motion, these limitations 
should be removed, and the HVT is make possible to remove 
limitation for the camera. Furthermore, human have to auto 
track by the HVT and any markers should be removed on the 
body. 
With this motive, marker less auto tracking method using the 
HVT was developed. 
Figure 3shows the basic flow of this auto tracking method, and 
detail auto tracking procedures are as follows. 
1. Pre-processing 
+ The search space in the tracking image which are 
taken by CCD 3 is specified. 
+ Gray level for the background in the specified area 
is acquired. 
+ The typical gray level of the subject is given, and 
the area for the target in the search space is 
calculated. 
+ Camera parameters (focal length, scale factor) for 
the CCD3 are inputted. 
2. Tracking image 
+ The HVT System is rotated to the start position and 
image for the subject on the start position is taken. 
+ Subject start walking. 
+ Tracking image is taken. 
3. Image processing 
+ The typical gray level of the subject is searched in 
the search space on tracking image. 
+ Extracted subject area is calculated. 
4. Image coordinate 
+ The central coordinates (center of gravity) of the 
extracted area are calculated. 
5. H.V. Angle 
+ Changing values for horizontal and vertical angles 
are calculated from image coordinate and camera 
parameters. 
6. Repeat the tracking procedure. 
In order to increasing the tracking speed, the resolution of 
tracking image was compressed as150 Pixel (H.)x120 Pixel 
(V.) (8 bit monochrome) in this paper. 
4. GAIT ANALYSIS 
4.1 Automatic Feature Extraction 
In order to evaluate difference in level at crossing. Gait analysis 
for the lower body, particularly ankle, heel and tiptoe is selected 
in this paper. 
The subjects extraction can be performed effectively using the 
remarkable feature of the HVT, and automatic extraction of 
-271- 
  
BENE MI a ae