Source-Channel Prediction for Error Resilient Video Coding
No Thumbnail Available
Date
2005-09
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Addis Ababa University
Abstract
Over recent decades the role of image in the communication of information has
grown steadly. Advances in technologies underlying the capture, transfer, storage, and
display of images as a means of communicating information has become technologically
and economically feasible. More importantly, images are in many situations an extremely
efficient means to communicate information.
The approach in this thesis work is to modify conventional motion compensation
prediction so as to improve or optimize the error resilience performance of the over all
system; motion compensated prediction in conventional video coders is based on the
encoder reconstruction of the previous frame where motion is estimated so as to
minimize the encoder prediction error. Such design paradigm optimizes the prediction at
the encoder and ignores the effect of packet loss.
The thesis work is based on the literature surveys, the simulation results are
based on the H.263 software and the necessary computation is done on MATLAB. The
modified (source-channel prediction) scheme achieves better error resilience performance
and better over all R-D trade off than its conventional counter part.
Description
Keywords
Video, Coding