(b) 
  
(d) 
  
(e) 
Figure 10: (a)Ziplock Snake (10x10) Gaussian Kernel Filter 
(b)Ziplock Snake (20x20) Gaussian Kernel Filter (c)Ziplock 
Snake (30x30) Gaussian Kernel Filter (d)Ziplock Snake 
(40x40) Gaussian Kernel Filter (¢)Ziplock Snake (50x50) 
Gaussian Kernel Filter 
In addition, optimal Gaussian Kernel Filter size depends on the 
resolution of the image. According to the experiments, if 
resolution of an image is changed, size of the Gaussian Kernel 
Filter must be changed by approximately the same ratio. Figure 
10(c) and 15 are converted from 1-meter resolution images to 2- 
meter resolution images using pyramidal decomposition. After 
that, Gaussian Kernel Filter in size of (15x15) and (10x10) is 
applied as pre-process for these low resolution images. Figure 
11 and 16 show results of road extraction algorithm. Depending 
on these figures, table 4 and table 5 are generated. According to 
these results, if resolution is decrease or increase, size of 
Gaussian kernel filter might be changed as same ratio with 
resolution change. 
   
(a) (b) 
Figure 11: (a) Ziplock Snake with (15x15) Gaussian Kernel 
Filter (b) Ziplock Snake with (10x10) Gaussian Kernel Filter 
  
   
  
(30x30) (15x15) (10x10) 
Figure 10(c) Figure 11(8) Figure 110). 
Correctness 9588.57 9485.53 7463.13 
  
  
i Gaussian Gaussian Gaussian 
3 Kernel Filter Kernel Filter Kernel Filter 
i Completeness 9697.74 9597.77 998.11 
  
  
  
  
| Image Size (pixel) 347x435 173x217 173x217 | 
  
Table 4: Evaluation of Different Size of Gaussian Kernel Filter 
for image in figure 10(c) and down sampled version in figure 11 
  
  
  
  
x Gaussian Gaussian Gaussian ; 
s Kernel Filter Kernel Filter Kernel Filter Ç 
(30x30) (15x15) (10x10) i 
i Figure 15 Figure 16(a) Figure 16(b) | 
| Correctness 955.83 %57.24 %47.53 | 
| Completeness %100 %100 %57.33 | 
i Image Size (pixel) 513x437 257x218 257x218 | 
m 
  
  
  
  
Table 5: Evaluation of Different Size of Gaussian Kernel Filter 
for image in figure 15 and down sampled version in figure 16 
3.2.2 Initialization 
The next step after the pre-processing is initialization. This step 
is performed by using Bezier Curves as shown in Figure 12. 
Bezier Curves are generated between the end points. The end 
points must have minimum gradient value. 
  
Figure 12: Bezier Curves 
Extended-snake based approach is developed by 
Neuenschwander et al. (1997) to prevent initialization and 
optimization problems. To achieve initialization procedure, 
homogeneous Euler equation is solved that it corresponds to 
system equation. Homogeneous Euler equation is defined as 
cn 
+ 
€ 
— 
La 
Cu 
ds 
a 
em 
Ly 
vs) 
where stands for either x or v and 0 < 5 < 1. 
According to the experiments, Ziplock snake detects straight 
roads, while it is not successful in detecting curved roads as 
shown in Figure 13. Ziplock snake needs some new control 
points that are defined by user during optimization. For an 
automatic approach, this situation is not appropriate. Thus, 
Ziplock snake is not capable of curved road detection 
automatically. As a consequence, if images that have straight 
roads are chosen, success of Ziplock snake method increases for 
automatic approaches. 
138 
Fi 
  
Fi