|
Activity Number:
|
517
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 5, 2009 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Statistics in Defense and National Security
|
| Abstract - #305749 |
|
Title:
|
Ternary Pixel Transition Matrix--Based Feature Space Statistics for Image Retrieval from Databases
|
|
Author(s):
|
Pranab K. Banerjee*+
|
|
Companies:
|
Space Dynamics Laboratory
|
|
Address:
|
1695 North Research Park Way, North Logan, UT, 84341,
|
|
Keywords:
|
Pixel transition ; Image retrieval ; Image processing ; Image similarity metric ; Feature extraction
|
|
Abstract:
|
Large digital image databases are common today in various application domains---from national defense to amateur photography. Manual image retrieval is impractical due to enormous data volume. Even semi-manual methods, such as manual image labeling followed by label space search, may be infeasible because of high volume and capture velocity. Thus, automated image retrieval is a necessity. This paper presents a statistical technique for automated search space reduction as well as automated retrieval of images based on feature space statistics. The feature extracted is based on ternary valued "pixel transition matrix" which is suitable for compact representation and efficient storage in the database. Automated search space reduction is key to enhancing image retrieval efficiency. It allows time efficient application of multiple retrieval algorithms on a database for robustness.
|
