國立虎尾科技大學 |

Image-Based Localization for Mobile and Vehicular Applications.

紀錄類型:	書目-語言資料,手稿 : Monograph/item
正題名/作者:	Image-Based Localization for Mobile and Vehicular Applications./
作者:	Sadeghi, Hamed.
面頁冊數:	1 online resource (156 pages)
附註:	Source: Dissertation Abstracts International, Volume: 79-04(E), Section: B.
Contained By:	Dissertation Abstracts International79-04B(E).
標題:	Electrical engineering. -
電子資源:	click for full text (PQDT)
ISBN:	9780355455076

Image-Based Localization for Mobile and Vehicular Applications.
Sadeghi, Hamed.

Image-Based Localization for Mobile and Vehicular Applications. - 1 online resource (156 pages)

Source: Dissertation Abstracts International, Volume: 79-04(E), Section: B.

Thesis (Ph.D.)

Includes bibliographical references

We study the image-based localization problem in mobile and vehicular applications. The problem can be stated as 'given a query image at an unknown location with some side information available regarding the camera and environment, how can one estimate the camera location with respect to a reference coordinates frame?'. The side information could be the query (or database) camera parameters, database of location-tagged images and/or measurements of some 3D dimensions. The problem has been well investigated in a number of literature works proposing methods for indoor and outdoor applications. We focus on the localization accuracy and complexity in this thesis and propose methods to improve them. Some of the proposed methods increase the location recognition rate, while others improve the (fine) localization accuracy. We also propose a robust and accurate method for outdoor scenarios.

Electronic reproduction.
Ann Arbor, Mich. :
ProQuest,
2018

Mode of access: World Wide Web

ISBN: 9780355455076Subjects--Topical Terms:

596380
Electrical engineering.
Index Terms--Genre/Form:

554714
Electronic books.

Image-Based Localization for Mobile and Vehicular Applications.
LDR:03864ntm a2200397Ki 4500 001 910718
005 20180517112608.5
006 m o u
007 cr mn||||a|a||
008 190606s2017 xx obm 000 0 eng d
020 $a 9780355455076
035 $a (MiAaPQ)AAI10256765
035 $a (MiAaPQ)toronto:15383
035 $a AAI10256765
040 $a MiAaPQ $b eng $c MiAaPQ
099 $a TUL $f hyy $c available through World Wide Web
100 1 $a Sadeghi, Hamed. $3 1182153
245 1 0 $a Image-Based Localization for Mobile and Vehicular Applications.
264 0 $c 2017
300 $a 1 online resource (156 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Dissertation Abstracts International, Volume: 79-04(E), Section: B.
500 $a Adviser: Shahrokh Valaee.
502 $a Thesis (Ph.D.) $c University of Toronto (Canada) $d 2017.
504 $a Includes bibliographical references
520 $a We study the image-based localization problem in mobile and vehicular applications. The problem can be stated as 'given a query image at an unknown location with some side information available regarding the camera and environment, how can one estimate the camera location with respect to a reference coordinates frame?'. The side information could be the query (or database) camera parameters, database of location-tagged images and/or measurements of some 3D dimensions. The problem has been well investigated in a number of literature works proposing methods for indoor and outdoor applications. We focus on the localization accuracy and complexity in this thesis and propose methods to improve them. Some of the proposed methods increase the location recognition rate, while others improve the (fine) localization accuracy. We also propose a robust and accurate method for outdoor scenarios.
520 $a The majority of the image-based localization methods can be classified into two categories, i.e. landmark-based (e.g. logo-based) and image retrieval-based methods. We propose a semi-supervised formulation for the logo-based problem, which results in a considerable decrease in the median and variance of the localization error compared to the state-of-the-art, at the cost of adding controllable computational complexity.
520 $a For the image retrieval-based approach, we propose an Epipolar geometry-based method to refine the localization estimate. The proposed method demonstrates greater accuracy, higher robustness to database locations sparsity as well as demonstrating extrapolation capability.
520 $a For the indoor scenarios with no textured landmarks and/or repeated scenery, where both category of methods fail to localize the user accurately, we propose two localization methods utilizing OCR, called OCRAPOSE I and OCRAPOSE II. The proposed methods improve the localization accuracy, provide simplified training (i.e. side information acquisition) and lower localization complexity.
520 $a For outdoor applications, we propose a two-dimensional robust localization method using the Google street view database, called 2DTriPnP. 2DTriPnP simplifies the complicated full 3D pose estimation by converting it to a 2D practical one and proposing a fast robust estimation method. It will be demonstrated in experiments that 2DTriPnP provides the great accuracy required in a number of outdoor scenarios such as safety in vehicular networks. Furthermore, it outperforms the 3D version in terms of the localization accuracy and complexity.
533 $a Electronic reproduction. $b Ann Arbor, Mich. : $c ProQuest, $d 2018
538 $a Mode of access: World Wide Web
650 4 $a Electrical engineering. $3 596380
650 4 $a Computer science. $3 573171
655 7 $a Electronic books. $2 local $3 554714
690 $a 0544
690 $a 0984
710 2 $a ProQuest Information and Learning Co. $3 1178819
710 2 $a University of Toronto (Canada). $b Electrical and Computer Engineering. $3 1148628
773 0 $t Dissertation Abstracts International $g 79-04B(E).
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=10256765 $z click for full text (PQDT)