Feature selection for position estimation using an omnidirectional camera

Huan N. Do, Mahdi Jadaliha, Jongeun Choi, Chae Young Lim

Research output: Contribution to journalArticle

14 Citations (Scopus)

Abstract

This paper considers visual feature selection to implement position estimation using an omnidirectional camera. The localization is based on a maximum likelihood estimation (MLE) with a map from optimally selected visual features using Gaussian process (GP) regression. In particular, the collection of selected features over a surveillance region is modeled by a multivariate GP with unknown hyperparameters. The hyperparameters are identified through the learning process by an MLE, which are used for prediction in an empirical Bayes fashion. To select features, we apply a backward sequential elimination technique in order to improve the quality of the position estimation with compressed features for efficient localization. The excellent results of the proposed algorithm are illustrated by the experimental studies with different visual features under both indoor and outdoor real-world scenarios.

Original languageEnglish
Pages (from-to)1-9
Number of pages9
JournalImage and Vision Computing
Volume39
DOIs
Publication statusPublished - 2015 Jul 1

Fingerprint

Maximum likelihood estimation
Feature extraction
Cameras

All Science Journal Classification (ASJC) codes

  • Signal Processing
  • Computer Vision and Pattern Recognition

Cite this

Do, Huan N. ; Jadaliha, Mahdi ; Choi, Jongeun ; Lim, Chae Young. / Feature selection for position estimation using an omnidirectional camera. In: Image and Vision Computing. 2015 ; Vol. 39. pp. 1-9.
@article{db495a6b021d43fba7b789749ca7c4ea,
title = "Feature selection for position estimation using an omnidirectional camera",
abstract = "This paper considers visual feature selection to implement position estimation using an omnidirectional camera. The localization is based on a maximum likelihood estimation (MLE) with a map from optimally selected visual features using Gaussian process (GP) regression. In particular, the collection of selected features over a surveillance region is modeled by a multivariate GP with unknown hyperparameters. The hyperparameters are identified through the learning process by an MLE, which are used for prediction in an empirical Bayes fashion. To select features, we apply a backward sequential elimination technique in order to improve the quality of the position estimation with compressed features for efficient localization. The excellent results of the proposed algorithm are illustrated by the experimental studies with different visual features under both indoor and outdoor real-world scenarios.",
author = "Do, {Huan N.} and Mahdi Jadaliha and Jongeun Choi and Lim, {Chae Young}",
year = "2015",
month = "7",
day = "1",
doi = "10.1016/j.imavis.2015.04.002",
language = "English",
volume = "39",
pages = "1--9",
journal = "Image and Vision Computing",
issn = "0262-8856",
publisher = "Elsevier Limited",

}

Feature selection for position estimation using an omnidirectional camera. / Do, Huan N.; Jadaliha, Mahdi; Choi, Jongeun; Lim, Chae Young.

In: Image and Vision Computing, Vol. 39, 01.07.2015, p. 1-9.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Feature selection for position estimation using an omnidirectional camera

AU - Do, Huan N.

AU - Jadaliha, Mahdi

AU - Choi, Jongeun

AU - Lim, Chae Young

PY - 2015/7/1

Y1 - 2015/7/1

N2 - This paper considers visual feature selection to implement position estimation using an omnidirectional camera. The localization is based on a maximum likelihood estimation (MLE) with a map from optimally selected visual features using Gaussian process (GP) regression. In particular, the collection of selected features over a surveillance region is modeled by a multivariate GP with unknown hyperparameters. The hyperparameters are identified through the learning process by an MLE, which are used for prediction in an empirical Bayes fashion. To select features, we apply a backward sequential elimination technique in order to improve the quality of the position estimation with compressed features for efficient localization. The excellent results of the proposed algorithm are illustrated by the experimental studies with different visual features under both indoor and outdoor real-world scenarios.

AB - This paper considers visual feature selection to implement position estimation using an omnidirectional camera. The localization is based on a maximum likelihood estimation (MLE) with a map from optimally selected visual features using Gaussian process (GP) regression. In particular, the collection of selected features over a surveillance region is modeled by a multivariate GP with unknown hyperparameters. The hyperparameters are identified through the learning process by an MLE, which are used for prediction in an empirical Bayes fashion. To select features, we apply a backward sequential elimination technique in order to improve the quality of the position estimation with compressed features for efficient localization. The excellent results of the proposed algorithm are illustrated by the experimental studies with different visual features under both indoor and outdoor real-world scenarios.

UR - http://www.scopus.com/inward/record.url?scp=84929621079&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84929621079&partnerID=8YFLogxK

U2 - 10.1016/j.imavis.2015.04.002

DO - 10.1016/j.imavis.2015.04.002

M3 - Article

AN - SCOPUS:84929621079

VL - 39

SP - 1

EP - 9

JO - Image and Vision Computing

JF - Image and Vision Computing

SN - 0262-8856

ER -