Home

About

Advanced Search

Browse by Discipline

Scientific Societies

E-print Alerts

Add E-prints

E-print Network
FAQHELPSITE MAPCONTACT US


  Advanced Search  

 
Modelling Objects using Distribution and Topology of Multiscale Region Pairs Himanshu Arora, Narendra Ahuja
 

Summary: Modelling Objects using Distribution and Topology of Multiscale Region Pairs
Himanshu Arora, Narendra Ahuja
Beckman Institute, University of Illinois at Urbana Champaign
Urbana, IL 61801, USA
{harora1,n-ahuja}@uiuc.edu
Abstract
We propose a method for simultaneous detection, local-
ization and segmentation of objects of a known category.
We show that this is possible by using segments as features.
To this end, we propose an object model in which the im-
age is represented as a tree, that captures containment re-
lationships among the segments. Using segments as fea-
tures has the advantage that object detection and segmen-
tation is done simultaneously, forgoing the need for a sep-
arate sophisticated model for object segmentation. A gen-
erative model of an object category is estimated in a super-
vised mode, in terms of the characteristics of its constituent
regions, their relative locations, and their mutual contain-
ment. The novel aspect of this work lies in simplifying the
description of the hierarchy in terms of constraints that ap-

  

Source: Ahuja, Narendra - Department of Electrical and Computer Engineering, University of Illinois at Urbana-Champaign

 

Collections: Computer Technologies and Information Sciences; Engineering