ICIP 2006, Atlanta, GA
 

Slide Show

Atlanta Conv. & Vis. Bureau

 

My ICIP 2006 Schedule

Note: Your custom schedule will not be saved unless you create a new account or login to an existing account.

Paper Detail

Paper:WA-P1.5
Session:Image and Video Segmentation
Time:Wednesday, October 11, 09:40 - 12:20
Presentation: Poster
Topic: Image & Video Segmentation: Statistical methods
Title: MAP-BASED OBJECT EXTRACTION FROM UNCALIBRATED IMAGE PAIR
Authors: Hyung Il Koo; Seoul National University 
 Sang Hwa Lee; Seoul National University 
 Nam Ik Cho; Seoul National University 
 Seong Keun Kim; SK Telecom 
 Dong Hahk Lee; SK Telecom 
 Sunghoon Lee; SK Telecom 
Abstract: In this paper, we propose a new MAP algorithm for foreground objects extraction from the uncalibrated image pair. The segmentation is modelled as a Markov random field and performed by the MAP framework. The proposed algorithm estimates several spatial transformations between two images by corresponding SIFT points and sequential RANSAC algorithm. And the area-ratio criterion is applied to the each transformation so that we select the transformation of foreground object. Using these transformations, we compute the likelihood of color-segmented subregions. We model the prior information which is based on minimum boundary length and smoothness. Finally, the object extraction is performed by the Bayesian belief propagation. Experimental results on various data sets show good performance to extract the foreground objects from the image pair.