|
During the past decades, the digital video data used by
computer applicationshave grown rapidly and enormously. As the
multimedia technologies become moreadvanced, there are more and
more digital video resources available. However,the widespread
uses of video in computer applications are often impeded bythe
lack of effective tools to manage video information. Most of
applicationsare toward how to compress/de-compress and playback
of video data. The notable features of video data are enormous
and unstructured. Videodatum itself is raw datum, and is created
independently from its contents.That is, we can't make any query
on raw video data. As the computer technologiesbecome more
advanced and more video resources become available, it is
importantthat we can enable both the manual and automated
retrieval/presentation videoinformation to meet the need of
content-based access.In this thesis, an integrated framework for
news processing was provided. Theauthor modified some
disadvantages of algorithm, and design a prototype systembased
on this modified model. The prototype system was programmed by
objectoriented methodology. By using this system, a half of
month of video data wasbe reviewed. The author tries to make
some conclusions based on these resultsand try to explain how
does these mean.In this thesis research, we show the ability of
personal computer to processnews video. There were two major
problems in this research, one is computer'sprocessing speed,
and the other is storage capacity. As the increase of
computerpower in processing video, these problems will be
conquered.
|