A Probabilistic Feature Map-Based Localization System Using a Monocular Camera
Image-based localization is one of the most widely researched localization techniques in the robotics and computer vision communities. As enormous image data sets are provided through the Internet, many studies on estimating a location with a pre-built image-based 3D map have been conducted. Most re...
Main Authors: | , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2015-08-01
|
Series: | Sensors |
Subjects: | |
Online Access: | http://www.mdpi.com/1424-8220/15/9/21636 |
id |
doaj-f47da676e07748599548d12841d1f06c |
---|---|
record_format |
Article |
spelling |
doaj-f47da676e07748599548d12841d1f06c2020-11-24T21:09:45ZengMDPI AGSensors1424-82202015-08-01159216362165910.3390/s150921636s150921636A Probabilistic Feature Map-Based Localization System Using a Monocular CameraHyungjin Kim0Donghwa Lee1Taekjun Oh2Hyun-Taek Choi3Hyun Myung4Urban Robotics Laboratory (URL), Korea Advanced Institute of Science and Technology (KAIST), 291 Daehak-ro (373-1 Guseong-dong), Yuseong-gu, Daejeon 305-701, KoreaUrban Robotics Laboratory (URL), Korea Advanced Institute of Science and Technology (KAIST), 291 Daehak-ro (373-1 Guseong-dong), Yuseong-gu, Daejeon 305-701, KoreaUrban Robotics Laboratory (URL), Korea Advanced Institute of Science and Technology (KAIST), 291 Daehak-ro (373-1 Guseong-dong), Yuseong-gu, Daejeon 305-701, KoreaOcean System Engineering Research Division, Korea Research Institute of Ships and Ocean Engineering (KRISO), 32 1312 Beon-gil, Yuseong-daero, Yuseong-gu, Daejeon 305-343, KoreaUrban Robotics Laboratory (URL), Korea Advanced Institute of Science and Technology (KAIST), 291 Daehak-ro (373-1 Guseong-dong), Yuseong-gu, Daejeon 305-701, KoreaImage-based localization is one of the most widely researched localization techniques in the robotics and computer vision communities. As enormous image data sets are provided through the Internet, many studies on estimating a location with a pre-built image-based 3D map have been conducted. Most research groups use numerous image data sets that contain sufficient features. In contrast, this paper focuses on image-based localization in the case of insufficient images and features. A more accurate localization method is proposed based on a probabilistic map using 3D-to-2D matching correspondences between a map and a query image. The probabilistic feature map is generated in advance by probabilistic modeling of the sensor system as well as the uncertainties of camera poses. Using the conventional PnP algorithm, an initial camera pose is estimated on the probabilistic feature map. The proposed algorithm is optimized from the initial pose by minimizing Mahalanobis distance errors between features from the query image and the map to improve accuracy. To verify that the localization accuracy is improved, the proposed algorithm is compared with the conventional algorithm in a simulation and realenvironmentshttp://www.mdpi.com/1424-8220/15/9/21636localizationmonocular cameraprobabilistic feature map3D-to-2D matching correspondencesimage data set |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Hyungjin Kim Donghwa Lee Taekjun Oh Hyun-Taek Choi Hyun Myung |
spellingShingle |
Hyungjin Kim Donghwa Lee Taekjun Oh Hyun-Taek Choi Hyun Myung A Probabilistic Feature Map-Based Localization System Using a Monocular Camera Sensors localization monocular camera probabilistic feature map 3D-to-2D matching correspondences image data set |
author_facet |
Hyungjin Kim Donghwa Lee Taekjun Oh Hyun-Taek Choi Hyun Myung |
author_sort |
Hyungjin Kim |
title |
A Probabilistic Feature Map-Based Localization System Using a Monocular Camera |
title_short |
A Probabilistic Feature Map-Based Localization System Using a Monocular Camera |
title_full |
A Probabilistic Feature Map-Based Localization System Using a Monocular Camera |
title_fullStr |
A Probabilistic Feature Map-Based Localization System Using a Monocular Camera |
title_full_unstemmed |
A Probabilistic Feature Map-Based Localization System Using a Monocular Camera |
title_sort |
probabilistic feature map-based localization system using a monocular camera |
publisher |
MDPI AG |
series |
Sensors |
issn |
1424-8220 |
publishDate |
2015-08-01 |
description |
Image-based localization is one of the most widely researched localization techniques in the robotics and computer vision communities. As enormous image data sets are provided through the Internet, many studies on estimating a location with a pre-built image-based 3D map have been conducted. Most research groups use numerous image data sets that contain sufficient features. In contrast, this paper focuses on image-based localization in the case of insufficient images and features. A more accurate localization method is proposed based on a probabilistic map using 3D-to-2D matching correspondences between a map and a query image. The probabilistic feature map is generated in advance by probabilistic modeling of the sensor system as well as the uncertainties of camera poses. Using the conventional PnP algorithm, an initial camera pose is estimated on the probabilistic feature map. The proposed algorithm is optimized from the initial pose by minimizing Mahalanobis distance errors between features from the query image and the map to improve accuracy. To verify that the localization accuracy is improved, the proposed algorithm is compared with the conventional algorithm in a simulation and realenvironments |
topic |
localization monocular camera probabilistic feature map 3D-to-2D matching correspondences image data set |
url |
http://www.mdpi.com/1424-8220/15/9/21636 |
work_keys_str_mv |
AT hyungjinkim aprobabilisticfeaturemapbasedlocalizationsystemusingamonocularcamera AT donghwalee aprobabilisticfeaturemapbasedlocalizationsystemusingamonocularcamera AT taekjunoh aprobabilisticfeaturemapbasedlocalizationsystemusingamonocularcamera AT hyuntaekchoi aprobabilisticfeaturemapbasedlocalizationsystemusingamonocularcamera AT hyunmyung aprobabilisticfeaturemapbasedlocalizationsystemusingamonocularcamera AT hyungjinkim probabilisticfeaturemapbasedlocalizationsystemusingamonocularcamera AT donghwalee probabilisticfeaturemapbasedlocalizationsystemusingamonocularcamera AT taekjunoh probabilisticfeaturemapbasedlocalizationsystemusingamonocularcamera AT hyuntaekchoi probabilisticfeaturemapbasedlocalizationsystemusingamonocularcamera AT hyunmyung probabilisticfeaturemapbasedlocalizationsystemusingamonocularcamera |
_version_ |
1716757454638809088 |