background image

 

13  

Abstract  

 
 
 

The  object  of  this  thesis  is  computer  vision.  With  the  advent  of  high-­‐speed  

computers   and   cameras   with   high   resolution   and   definition,   the   research  
interest  was  expanded  for  the  development  and  implementation  of  algorithms  in  

the  field  of  computer  vision.  Algorithms  for  applications  vary  depending  on  their  
field   of   interest.   Applications   can   be   developed   with   computer   vision   for  

security,   military   and   intelligence,   for   health   issues,   research   and   production.  
Among   these   applications,   one   of   the   most   frequently   performed   task   is   to  
determine   the   position   and   the   motion   detection   for   each   subject,   all   of   which  

occupied  us  to  develop  our  application.  

 
Nowadays,  the  external  aesthetics  and  integrity  of  the  products  is  obtained  

in  an  equal  degree  with  regard  to  the  product  itself  from  the  general  public,  a  
fact   that   preoccupies   manufacturers.   To   ensure   long-­‐term   promotion   of   a  

product,   the   appearance   and   quality   product   integrity   must   be   visually  
inspected   before   dispatch.   This   desired   result   can   be   achieved   if   the   process  
inspection  become  automated  and  is  performed  without  human  intervention.  

 
By  using  the  techniques  offered  on  machine  vision  tasks,  we  designed  and  

implemented   an   application   for   product   classification   in   a   transmission   line  
using  camera,  i.e.  the  use  of  a  sequence  of  images.  This  application  enables  the  
holder  to  have  a  complete  picture  regarding  quality  and  quantity  of  the  products  

thus  avoiding  any  human  errors.  

 
At  this  point  I  would  like  to  thank  my  supervisor  Dr.  Mina  Dasygeni  for  his  

valuable  help  during  the  implementation  of  this  thesis  and  for  the  knowledge  I  
gained   thanks   to   him.   The   topic   of   the   thesis   was   proposed   by   Dr.   Minas  

Dasygenis  and  the  application  was  developed  with  programming  language  C++  
using  the  libraries  of  OpenCV.  
 

Keywords:  OpenCV,  C  +  +,  Computer  Vision,  Production  line.  
 

 
 
 

 
 
 

Kozani,  June  2014  

 

Loizou  Konstantina