Your browser does not support JavaScript!

Home    Search  

Results - Details

Search command : Author="Komodakis"  And Author="Nikos"

Current Record: 2 of 2

Back to Results Previous page
Next page
Add to Basket
[Add to Basket]
Identifier uch.csd.msc//1997komodakis
Title Εκτίμηση Τρισδιάστατης Κίνησης και Εύρεση Στρωμάτων Βάθους σε Ακολουθίες Εικόνων
Alternative Title 3-D Motion Estimation and Image Depth Layering
Creator Komodakis, Nikos
Abstract In this thesis two problems are addressed: estimation of camera 3-D motion and image depth layering. For the first problem, the method we adopt consists of two stages. During the first stage a 2-D motion field is computed. During the second stage, the 3-D motion parameters are identified by equations linking the projected 2-D motions and 3-D motions inside the image sequence. During this stage, the main problem comes from the fact that the optical flow field may contain a set of noisy and partially incorrect data (outliers). The negative effects of this set of outliers to the motion estimation, increase with the complexity of the motion model that is used to describe the camera motion. Therefore, the approach we adopt relies on the use of a hierarchy of motion models. Simpler models are first tested, and then more complex models are considered. In this way we aim at finding the simplest possible model, which can adequately describe the camera motion. Additionally, at each stage of the algorithm a robust estimation method is used to cope with the set of outliers. The second problem, that is image depth layering, is ultimately reduced to the problem of segmenting the optical flow field. For this purpose, an iterative deterministic relaxation algorithm is used and the extraction of the image depth layers is based on both optical flow and luminance information. The iterative algorithm employed in this thesis is the HCF (Highest Confidence First). Experimental results regarding real and synthetic image sequences are given to demonstrate the proposed methods.
Issue date 1997-11-01
Date available 1998-04-0
Collection   School/Department--School of Sciences and Engineering--Department of Computer Science--Post-graduate theses
  Type of Work--Post-graduate theses
Permanent Link Bookmark and Share
Views 276

Digital Documents
No preview available

View document

No preview available

Download document
View document
Views : 3