Vehicle Category Recognition from Aerial Videos

Ozge C. Ozcanli, Amir Tamrakar, Benjamin B. Kimia, Joseph L. Mundy

 

 

The aim of this project is to recognize generic categories of vehicles from aerial videos.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

We register the videos and use background modeling to segment the vehicles and form prototype databases.

 

 

 

 

 

 

 

 

 

 

We propose various algorithms to match these prototypes and determine the category.

A testbed of still image databases are generated to test prototype matching algorithms for recognition.

 

 

 

 

 

 

 

 

Browse the LEMS Vehicle (still images) database

Download the LEMS Vehicle database

 

 

Publications:

 

O. C. Ozcanli, A. Tamrakar, B. B. Kimia and J. L. Mundy, Augmenting Shape with Appearance in Vehicle Category Recognition, CVPR06, New York, 2006. pdf,

CVPR 06 Presentation

 

 

 

 

 

 

 

 

Last updated Aug 09, 2006

ozge @ lems.brown.edu

 

 

 

 

 

 

 

 

 

 

Last Updated: Aug 09, 2006

 

 

 

 

 

 

 

minivan

passenger car

passenger car

...

...

Database for view k

Database for view m