Difference between revisions of "Dp 332 en"

From DCEwiki
Jump to: navigation, search
Line 4: Line 4:
 
   
 
   
 
[[Diplomové práce 2009]]
 
[[Diplomové práce 2009]]
 +
 +
[[Image:332_dp.gif]]
  
 
The aim of this diploma thesis was to design an obstacle detection system for space
 
The aim of this diploma thesis was to design an obstacle detection system for space
Line 12: Line 14:
 
platform of development kit complemented by an expansion module with two CMOS
 
platform of development kit complemented by an expansion module with two CMOS
 
cameras.
 
cameras.
 +
 
Firstly, the algorithmization of a simple and fast stereovision algorithm was implemented
 
Firstly, the algorithmization of a simple and fast stereovision algorithm was implemented
 
in Matlab. Later on, it was applied to the chosen signal processor. An effective
 
in Matlab. Later on, it was applied to the chosen signal processor. An effective
Line 17: Line 20:
 
cameras is being carried out via rectification and calibration of a stereoscopic image pair.
 
cameras is being carried out via rectification and calibration of a stereoscopic image pair.
 
In conclusion, the work summarizes the results of implemented methods and possible
 
In conclusion, the work summarizes the results of implemented methods and possible
improvements to the proposal
+
improvements to the proposal.
 +
 
 +
* '''Bc. Ondřej Ton''', tel: +420 739 666 831, mailto:tono1@fel.cvut.cz

Revision as of 13:00, 5 May 2010

Orientation in space

Author: Ton Ondřej

Diplomové práce 2009

332 dp.gif

The aim of this diploma thesis was to design an obstacle detection system for space navigation focused on the mobile devices models developed at the Department of Control Engineering CTU. It summarized the basic methods of measuring the depth map of the scene, of which chosen method is stereovision. This method has been analysed and applied to the selected signal DSP processor Blackfin ADSP-BF61 using a specified hardware platform of development kit complemented by an expansion module with two CMOS cameras.

Firstly, the algorithmization of a simple and fast stereovision algorithm was implemented in Matlab. Later on, it was applied to the chosen signal processor. An effective correspondence search between stereoscopic images of outside world acquired by 2 CMOS cameras is being carried out via rectification and calibration of a stereoscopic image pair. In conclusion, the work summarizes the results of implemented methods and possible improvements to the proposal.