Loading…

Robust Rotation and Translation Estimation in Multiview Reconstruction

It is known that the problem of multiview reconstruction can be solved in two steps: first estimate camera rotations and then translations using them. This paper presents new robust techniques for both of these steps, (i) Given pair-wise relative rotations, global camera rotations are estimated line...

Full description

Saved in:
Bibliographic Details
Main Authors: Martinec, D., Pajdla, T.
Format: Conference Proceeding
Language:English
Subjects:
Citations: Items that cite this one
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
cited_by cdi_FETCH-LOGICAL-c2455-8226a7e2f51268cc87644febc47999d6e9cebffb2b4eb431fd5543e0bd68d85c3
cites
container_end_page 8
container_issue
container_start_page 1
container_title
container_volume
creator Martinec, D.
Pajdla, T.
description It is known that the problem of multiview reconstruction can be solved in two steps: first estimate camera rotations and then translations using them. This paper presents new robust techniques for both of these steps, (i) Given pair-wise relative rotations, global camera rotations are estimated linearly in least squares, (ii) Camera translations are estimated using a standard technique based on Second Order Cone Programming. Robustness is achieved by using only a subset of points according to a new criterion that diminishes the risk of choosing a mismatch. It is shown that only four points chosen in a special way are sufficient to represent a pairwise reconstruction almost equally as all points. This leads to a significant speedup. In image sets with repetitive or similar structures, non-existent epipolar geometries may be found. Due to them, some rotations and consequently translations may be estimated incorrectly. It is shown that iterative removal of pairwise reconstructions with the largest residual and reregistration removes most non-existent epipolar geometries. The performance of the proposed method is demonstrated on difficult wide base-line image sets.
doi_str_mv 10.1109/CVPR.2007.383115
format conference_proceeding
fullrecord <record><control><sourceid>ieee_6IE</sourceid><recordid>TN_cdi_ieee_primary_4270140</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><ieee_id>4270140</ieee_id><sourcerecordid>4270140</sourcerecordid><originalsourceid>FETCH-LOGICAL-c2455-8226a7e2f51268cc87644febc47999d6e9cebffb2b4eb431fd5543e0bd68d85c3</originalsourceid><addsrcrecordid>eNotTs1KxDAYjKjguvYueOkLtObLf45SdldhRSmr16VJE4jUVppU8e2tdOcyM8wwDEK3gEsArO-r99e6JBjLkioKwM_QNTDCGIDC8hxlWqqTl5pfoBVgQQuhQV-hLMYPPEPNVa5WaFsPZoopr4fUpDD0edO3-WFs-tgtfhNT-Fxk6PPnqUvhO7ifvHZ26GMaJ_uf3aBL33TRZSdeo7ft5lA9FvuX3VP1sC8sYZwXihDRSEc8ByKUtUoKxrwzlkmtdSucts54b4hhzjAKvuWcUYdNK1SruKVrdLfsBufc8Wucr42_R0YkBobpHz49T2g</addsrcrecordid><sourcetype>Publisher</sourcetype><iscdi>true</iscdi><recordtype>conference_proceeding</recordtype></control><display><type>conference_proceeding</type><title>Robust Rotation and Translation Estimation in Multiview Reconstruction</title><source>IEEE Electronic Library (IEL) Conference Proceedings</source><creator>Martinec, D. ; Pajdla, T.</creator><creatorcontrib>Martinec, D. ; Pajdla, T.</creatorcontrib><description>It is known that the problem of multiview reconstruction can be solved in two steps: first estimate camera rotations and then translations using them. This paper presents new robust techniques for both of these steps, (i) Given pair-wise relative rotations, global camera rotations are estimated linearly in least squares, (ii) Camera translations are estimated using a standard technique based on Second Order Cone Programming. Robustness is achieved by using only a subset of points according to a new criterion that diminishes the risk of choosing a mismatch. It is shown that only four points chosen in a special way are sufficient to represent a pairwise reconstruction almost equally as all points. This leads to a significant speedup. In image sets with repetitive or similar structures, non-existent epipolar geometries may be found. Due to them, some rotations and consequently translations may be estimated incorrectly. It is shown that iterative removal of pairwise reconstructions with the largest residual and reregistration removes most non-existent epipolar geometries. The performance of the proposed method is demonstrated on difficult wide base-line image sets.</description><identifier>ISSN: 1063-6919</identifier><identifier>ISBN: 9781424411795</identifier><identifier>ISBN: 1424411793</identifier><identifier>EISBN: 1424411807</identifier><identifier>EISBN: 9781424411801</identifier><identifier>DOI: 10.1109/CVPR.2007.383115</identifier><language>eng</language><publisher>IEEE</publisher><subject>Buildings ; Cameras ; Cybernetics ; Geometry ; Image reconstruction ; Jacobian matrices ; Layout ; Least squares approximation ; Linear programming ; Robustness</subject><ispartof>2007 IEEE Conference on Computer Vision and Pattern Recognition, 2007, p.1-8</ispartof><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c2455-8226a7e2f51268cc87644febc47999d6e9cebffb2b4eb431fd5543e0bd68d85c3</citedby></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://ieeexplore.ieee.org/document/4270140$$EHTML$$P50$$Gieee$$H</linktohtml><link.rule.ids>309,310,780,784,789,790,2058,27925,54920</link.rule.ids><linktorsrc>$$Uhttps://ieeexplore.ieee.org/document/4270140$$EView_record_in_IEEE$$FView_record_in_$$GIEEE</linktorsrc></links><search><creatorcontrib>Martinec, D.</creatorcontrib><creatorcontrib>Pajdla, T.</creatorcontrib><title>Robust Rotation and Translation Estimation in Multiview Reconstruction</title><title>2007 IEEE Conference on Computer Vision and Pattern Recognition</title><addtitle>CVPR</addtitle><description>It is known that the problem of multiview reconstruction can be solved in two steps: first estimate camera rotations and then translations using them. This paper presents new robust techniques for both of these steps, (i) Given pair-wise relative rotations, global camera rotations are estimated linearly in least squares, (ii) Camera translations are estimated using a standard technique based on Second Order Cone Programming. Robustness is achieved by using only a subset of points according to a new criterion that diminishes the risk of choosing a mismatch. It is shown that only four points chosen in a special way are sufficient to represent a pairwise reconstruction almost equally as all points. This leads to a significant speedup. In image sets with repetitive or similar structures, non-existent epipolar geometries may be found. Due to them, some rotations and consequently translations may be estimated incorrectly. It is shown that iterative removal of pairwise reconstructions with the largest residual and reregistration removes most non-existent epipolar geometries. The performance of the proposed method is demonstrated on difficult wide base-line image sets.</description><subject>Buildings</subject><subject>Cameras</subject><subject>Cybernetics</subject><subject>Geometry</subject><subject>Image reconstruction</subject><subject>Jacobian matrices</subject><subject>Layout</subject><subject>Least squares approximation</subject><subject>Linear programming</subject><subject>Robustness</subject><issn>1063-6919</issn><isbn>9781424411795</isbn><isbn>1424411793</isbn><isbn>1424411807</isbn><isbn>9781424411801</isbn><fulltext>true</fulltext><rsrctype>conference_proceeding</rsrctype><creationdate>2007</creationdate><recordtype>conference_proceeding</recordtype><sourceid>6IE</sourceid><recordid>eNotTs1KxDAYjKjguvYueOkLtObLf45SdldhRSmr16VJE4jUVppU8e2tdOcyM8wwDEK3gEsArO-r99e6JBjLkioKwM_QNTDCGIDC8hxlWqqTl5pfoBVgQQuhQV-hLMYPPEPNVa5WaFsPZoopr4fUpDD0edO3-WFs-tgtfhNT-Fxk6PPnqUvhO7ifvHZ26GMaJ_uf3aBL33TRZSdeo7ft5lA9FvuX3VP1sC8sYZwXihDRSEc8ByKUtUoKxrwzlkmtdSucts54b4hhzjAKvuWcUYdNK1SruKVrdLfsBufc8Wucr42_R0YkBobpHz49T2g</recordid><startdate>200706</startdate><enddate>200706</enddate><creator>Martinec, D.</creator><creator>Pajdla, T.</creator><general>IEEE</general><scope>6IE</scope><scope>6IH</scope><scope>CBEJK</scope><scope>RIE</scope><scope>RIO</scope></search><sort><creationdate>200706</creationdate><title>Robust Rotation and Translation Estimation in Multiview Reconstruction</title><author>Martinec, D. ; Pajdla, T.</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c2455-8226a7e2f51268cc87644febc47999d6e9cebffb2b4eb431fd5543e0bd68d85c3</frbrgroupid><rsrctype>conference_proceedings</rsrctype><prefilter>conference_proceedings</prefilter><language>eng</language><creationdate>2007</creationdate><topic>Buildings</topic><topic>Cameras</topic><topic>Cybernetics</topic><topic>Geometry</topic><topic>Image reconstruction</topic><topic>Jacobian matrices</topic><topic>Layout</topic><topic>Least squares approximation</topic><topic>Linear programming</topic><topic>Robustness</topic><toplevel>online_resources</toplevel><creatorcontrib>Martinec, D.</creatorcontrib><creatorcontrib>Pajdla, T.</creatorcontrib><collection>IEEE Electronic Library (IEL) Conference Proceedings</collection><collection>IEEE Proceedings Order Plan (POP) 1998-present by volume</collection><collection>IEEE Xplore All Conference Proceedings</collection><collection>IEEE Electronic Library (IEL)</collection><collection>IEEE Proceedings Order Plans (POP) 1998-present</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Martinec, D.</au><au>Pajdla, T.</au><format>book</format><genre>proceeding</genre><ristype>CONF</ristype><atitle>Robust Rotation and Translation Estimation in Multiview Reconstruction</atitle><btitle>2007 IEEE Conference on Computer Vision and Pattern Recognition</btitle><stitle>CVPR</stitle><date>2007-06</date><risdate>2007</risdate><spage>1</spage><epage>8</epage><pages>1-8</pages><issn>1063-6919</issn><isbn>9781424411795</isbn><isbn>1424411793</isbn><eisbn>1424411807</eisbn><eisbn>9781424411801</eisbn><abstract>It is known that the problem of multiview reconstruction can be solved in two steps: first estimate camera rotations and then translations using them. This paper presents new robust techniques for both of these steps, (i) Given pair-wise relative rotations, global camera rotations are estimated linearly in least squares, (ii) Camera translations are estimated using a standard technique based on Second Order Cone Programming. Robustness is achieved by using only a subset of points according to a new criterion that diminishes the risk of choosing a mismatch. It is shown that only four points chosen in a special way are sufficient to represent a pairwise reconstruction almost equally as all points. This leads to a significant speedup. In image sets with repetitive or similar structures, non-existent epipolar geometries may be found. Due to them, some rotations and consequently translations may be estimated incorrectly. It is shown that iterative removal of pairwise reconstructions with the largest residual and reregistration removes most non-existent epipolar geometries. The performance of the proposed method is demonstrated on difficult wide base-line image sets.</abstract><pub>IEEE</pub><doi>10.1109/CVPR.2007.383115</doi><tpages>8</tpages><oa>free_for_read</oa></addata></record>
fulltext fulltext_linktorsrc
identifier ISSN: 1063-6919
ispartof 2007 IEEE Conference on Computer Vision and Pattern Recognition, 2007, p.1-8
issn 1063-6919
language eng
recordid cdi_ieee_primary_4270140
source IEEE Electronic Library (IEL) Conference Proceedings
subjects Buildings
Cameras
Cybernetics
Geometry
Image reconstruction
Jacobian matrices
Layout
Least squares approximation
Linear programming
Robustness
title Robust Rotation and Translation Estimation in Multiview Reconstruction
url http://sfxeu10.hosted.exlibrisgroup.com/loughborough?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-25T02%3A40%3A15IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-ieee_6IE&rft_val_fmt=info:ofi/fmt:kev:mtx:book&rft.genre=proceeding&rft.atitle=Robust%20Rotation%20and%20Translation%20Estimation%20in%20Multiview%20Reconstruction&rft.btitle=2007%20IEEE%20Conference%20on%20Computer%20Vision%20and%20Pattern%20Recognition&rft.au=Martinec,%20D.&rft.date=2007-06&rft.spage=1&rft.epage=8&rft.pages=1-8&rft.issn=1063-6919&rft.isbn=9781424411795&rft.isbn_list=1424411793&rft_id=info:doi/10.1109/CVPR.2007.383115&rft.eisbn=1424411807&rft.eisbn_list=9781424411801&rft_dat=%3Cieee_6IE%3E4270140%3C/ieee_6IE%3E%3Cgrp_id%3Ecdi_FETCH-LOGICAL-c2455-8226a7e2f51268cc87644febc47999d6e9cebffb2b4eb431fd5543e0bd68d85c3%3C/grp_id%3E%3Coa%3E%3C/oa%3E%3Curl%3E%3C/url%3E&rft_id=info:oai/&rft_id=info:pmid/&rft_ieee_id=4270140&rfr_iscdi=true