Stereo registration using kernel density correlation

Maneesh Singh, Ashish Jagmohan, Narendra Ahuja

Research output: Contribution to conferencePaper

Abstract

A common approach to solving the stereo registration problem is to model the disparity function as a discrete-valued Markov Random Field. The key problems with this approach are its combinatorial computational complexity, and the discretization of the obtained disparity estimates. In this paper, we propose a framework that addresses the requirements of a robust continuous domain formulation for stereo registration. The proposed formulation is based on a new measure, derived from the correlation of empirical probability density distributions estimated using kernel estimators. We term this the kernel density correlation (KDC) measure. The proposed framework takes the form of an energy minimization formulation which is efficiently solved using the technique of variational optimization. We prove the convergence properties of the resultant iterative algorithm, and compare the performance of the proposed formulation to that of a state-of-the-art stereo registration approach.

Original languageEnglish (US)
Pages260-265
Number of pages6
StatePublished - Dec 27 2004
EventProceedings of the Seventh IASTED International Conference on Computer Graphics and Imaging - Kauai, HI, United States
Duration: Aug 17 2004Aug 19 2004

Other

OtherProceedings of the Seventh IASTED International Conference on Computer Graphics and Imaging
CountryUnited States
CityKauai, HI
Period8/17/048/19/04

Fingerprint

Computational complexity

Keywords

  • Dense disparity field
  • Density estimation
  • MRF
  • Parzen windows
  • Stereo
  • Variational optimization

ASJC Scopus subject areas

  • Engineering(all)

Cite this

Singh, M., Jagmohan, A., & Ahuja, N. (2004). Stereo registration using kernel density correlation. 260-265. Paper presented at Proceedings of the Seventh IASTED International Conference on Computer Graphics and Imaging, Kauai, HI, United States.

Stereo registration using kernel density correlation. / Singh, Maneesh; Jagmohan, Ashish; Ahuja, Narendra.

2004. 260-265 Paper presented at Proceedings of the Seventh IASTED International Conference on Computer Graphics and Imaging, Kauai, HI, United States.

Research output: Contribution to conferencePaper

Singh, M, Jagmohan, A & Ahuja, N 2004, 'Stereo registration using kernel density correlation', Paper presented at Proceedings of the Seventh IASTED International Conference on Computer Graphics and Imaging, Kauai, HI, United States, 8/17/04 - 8/19/04 pp. 260-265.
Singh M, Jagmohan A, Ahuja N. Stereo registration using kernel density correlation. 2004. Paper presented at Proceedings of the Seventh IASTED International Conference on Computer Graphics and Imaging, Kauai, HI, United States.
Singh, Maneesh ; Jagmohan, Ashish ; Ahuja, Narendra. / Stereo registration using kernel density correlation. Paper presented at Proceedings of the Seventh IASTED International Conference on Computer Graphics and Imaging, Kauai, HI, United States.6 p.
@conference{19c8d686c57b4bef91c7cd50854e2d5e,
title = "Stereo registration using kernel density correlation",
abstract = "A common approach to solving the stereo registration problem is to model the disparity function as a discrete-valued Markov Random Field. The key problems with this approach are its combinatorial computational complexity, and the discretization of the obtained disparity estimates. In this paper, we propose a framework that addresses the requirements of a robust continuous domain formulation for stereo registration. The proposed formulation is based on a new measure, derived from the correlation of empirical probability density distributions estimated using kernel estimators. We term this the kernel density correlation (KDC) measure. The proposed framework takes the form of an energy minimization formulation which is efficiently solved using the technique of variational optimization. We prove the convergence properties of the resultant iterative algorithm, and compare the performance of the proposed formulation to that of a state-of-the-art stereo registration approach.",
keywords = "Dense disparity field, Density estimation, MRF, Parzen windows, Stereo, Variational optimization",
author = "Maneesh Singh and Ashish Jagmohan and Narendra Ahuja",
year = "2004",
month = "12",
day = "27",
language = "English (US)",
pages = "260--265",
note = "Proceedings of the Seventh IASTED International Conference on Computer Graphics and Imaging ; Conference date: 17-08-2004 Through 19-08-2004",

}

TY - CONF

T1 - Stereo registration using kernel density correlation

AU - Singh, Maneesh

AU - Jagmohan, Ashish

AU - Ahuja, Narendra

PY - 2004/12/27

Y1 - 2004/12/27

N2 - A common approach to solving the stereo registration problem is to model the disparity function as a discrete-valued Markov Random Field. The key problems with this approach are its combinatorial computational complexity, and the discretization of the obtained disparity estimates. In this paper, we propose a framework that addresses the requirements of a robust continuous domain formulation for stereo registration. The proposed formulation is based on a new measure, derived from the correlation of empirical probability density distributions estimated using kernel estimators. We term this the kernel density correlation (KDC) measure. The proposed framework takes the form of an energy minimization formulation which is efficiently solved using the technique of variational optimization. We prove the convergence properties of the resultant iterative algorithm, and compare the performance of the proposed formulation to that of a state-of-the-art stereo registration approach.

AB - A common approach to solving the stereo registration problem is to model the disparity function as a discrete-valued Markov Random Field. The key problems with this approach are its combinatorial computational complexity, and the discretization of the obtained disparity estimates. In this paper, we propose a framework that addresses the requirements of a robust continuous domain formulation for stereo registration. The proposed formulation is based on a new measure, derived from the correlation of empirical probability density distributions estimated using kernel estimators. We term this the kernel density correlation (KDC) measure. The proposed framework takes the form of an energy minimization formulation which is efficiently solved using the technique of variational optimization. We prove the convergence properties of the resultant iterative algorithm, and compare the performance of the proposed formulation to that of a state-of-the-art stereo registration approach.

KW - Dense disparity field

KW - Density estimation

KW - MRF

KW - Parzen windows

KW - Stereo

KW - Variational optimization

UR - http://www.scopus.com/inward/record.url?scp=10444265965&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=10444265965&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:10444265965

SP - 260

EP - 265

ER -