1. Image enhancement 
1-1 Appointment of window 
1-2 Smoothing with edge preserving 
1-3 Operation between different images 
to erase white track lines 
1-4 Sharpening by Laplacian 
-5 Binarization 
1-6 Preparation of Histogram for X and Y 
directions to erase small region of 
noises 
1-7 Region Segmentation 
In this step, small region can be 
erased based on the histogram which 
were prepared in stepi-6. 
2. Morphological Operation 
2-1 Expansion 
2-2 Contraction 
2-3 Thinning by Hilditch 
3. Feature Extraction 
3-1 Skeleton 
3-2 Feature Points 
Fig.2 shows an original 4.8 second image 
while Fig.3,4 and 5 show the results of 
each step image. Fig.6 shows overlapped 
image on the original image. Table 1 
shows the results of pixel and line 
number for the feature points(Fig.7) of 
Carl Lowis at 4.8 seconds from the start. 
Table 1 Pixel and Line 
  
of Feature point | 
12 
de NA 
No Pixel Line ji 
26 
1 289 201 
2 288 211 sd 
3 263 238 72 
4 308 227 N 
5 271. 218 a9 
6 297 218 10, ag 
7 288 241 {2 
8 281 259 
9 298 255 
10 267 259 1 
il: 30274 284 Wo 
12 267 262 
13 305 286 
T Fig.7 Feature Points 
3. Sports dynamics of Carl Lowis 
Recently sports dynamics analysis 
has received more attention. The basic 
physical elements of this analysis can be 
summarized as follows, 
1.Mechanical energy=potential energy(mgh) 
+ kinetic energy (mvZ/2) 
    
  
  
  
   
   
   
   
   
   
  
       
    
   
   
    
   
    
     
   
     
  
       
         
     
   
    
     
  
  
Fig.6 Compound Image 
pixel 
10r 
EAS 
—15! 
gu. Oo c 
~~ 
  
-20 past ae a a a a A 
  
4.0 4.1 4.2 4.3 4.4 4.5 4.6 
sec 
Fig.8 Horizontal Displacement of Right Knee 
2.Impulsive force=Newton's 2nd Law of 
Motion(ma) 
where m;mass weight(Carl louis  80kg), 
g;the acceleration of gravity, h;height, 
v;velocity, a;acceleration 
Fig.8 shows the results of the horizontal 
displacement of Carl Lowis's right knee. 
This graph shows the motion from right- 
foot push off(4.1 sec.) to the next right 
foot push  off(4.57  sec.). "This cycle 
represents one stride of running. 
The pixel(Vertical axis in Fig.8) 
represents the horizontal displacement 
from the pixel at the top of the head to 
that at the right knee. Values below zero 
thus mean that the right knee is located 
behind the top of the head, while values 
above zero mean that the right knee is 
ahead of the top of the head. At a pixel 
value of zero, the opposite leg is just 
crossing. The minimum pixel value at 4.2 
sec. thus mean that the knee is farthest 
behind from the head (the knee is most