more than 2,the point is non — isolated and non — structure to represent the image parallel thinning for 
end. The current point is a signal, as long as the the north border algorithm as provided in figure 1. 
value of this point is 1. Hence, we can use matrix 
  
Thinning 
Fig. 1 Block Diagram for Image Thinning 
The experimental result for this methodology is 
showed in figure 2. Fig. 2 a). The Original Image. 
5 E En Ideen b). The Thin—line Image after 
Performing Image Thinning 
  
  
  
3. BORDER FOLLOWING AND 
0 MAIRIX REPRESENTATION 
  
  
Border following is one of the underlying techniques 
in the processing of digitized binary images, and its 
techniques have also found applications in various 
other related problems, i. e. pattern recognition, 
image analysis and image data compression. In 
general, the borders of a subset S of an image are 
considered as the set of points ad jacent to S in —S 
a RS E is do which is the complement of S. The traditional 
approach to find borders is that at first find a starting 
point , decide the direction of the border in the 8 
neighbors , and follow the trend. Moreover, the 
USES | T intersection points must be put in the store, so that 
rali } those points can be found in next loop. Especially, 
12