|
The
act of cataloging video is a large task, the intent of the database
drives the management and indexing of the video. The basic operation
is matching during retrieval. A similarity metric is needed to
find matches, but how similarity is represented and stored varies
on the application. As the amount of video resources increases
it becomes more and more impracticable for manual annotation of
video to be the main device for retrieval. Therefore automatic
feature extraction is the usually means of comparison in large
video databases. One promising technique is to query with a sample
clip and return similar database entries. In this talk I will
present the results from examining the query-by-example technique
described in Dimitrova and Abdel-Mottaleb [1] and discuss refinements
that can be applied to this technique. [1] N. Dimitrova and M.
Abdel-Mottaleb. Content-based video retrieval by example video
clip. In Proceedings of the SPIE conference on Storage and Retrieval
for Image and Video Databases V, Volume 3022 , pages 59-70, 1996.
|
