1094 articles – 620 Notices  [english version]
HAL : hal-00647995, version 1

Fiche concise  Récupérer au format
Meaningful Matches in Stereovision
Sabater N., Almansa A., Morel J.-M.
IEEE Transactions on Pattern Analysis and Machine Intelligence 34, 5 (2012) 930-942 - http://hal.archives-ouvertes.fr/hal-00647995
Articles dans des revues avec comité de lecture
Informatique/Vision par ordinateur et reconnaissance de formes
Meaningful Matches in Stereovision
Neus Sabater () 1, Andrés Almansa () 2, Jean-Michel Morel 3
1 :  Centre de Mathématiques et de Leurs Applications - ENS Cachan (CMLA)
http://www.cmla.ens-cachan.fr/
École normale supérieure (ENS) - Cachan
ENS Cachan CNRS, PRES UniverSud, 61 Avenue du Pr´esident Wilson 94230 Cachan
France
2 :  Laboratoire Traitement et Communication de l'Information [Paris] (LTCI)
http://www.ltci.telecom-paristech.fr/
Télécom ParisTech – CNRS : UMR5141
CNRS LTCI Télécom ParisTech 46 rue Barrault F-75634 Paris Cedex 13
France
3 :  Centre de Mathématiques et de Leurs Applications (CMLA)
http://www.cmla.ens-cachan.fr/Cmla/index.html
CNRS : UMR8536 – École normale supérieure (ENS) - Cachan
France
This paper introduces a statistical method to decide whether two blocks in a pair of of images match reliably. The method ensures that the selected block matches are unlikely to have occurred "just by chance.'' The new approach is based on the definition of a simple but faithful statistical "background model" for image blocks learned from the image itself. A theorem guarantees that under this model not more than a fixed number of wrong matches occurs (on average) for the whole image. This fixed number (the number of false alarms) is the only method parameter. Furthermore, the number of false alarms associated with each match measures its reliability. This "a contrario" block-matching method, however, cannot rule out false matches due to the presence of periodic objects in the images. But it is successfully complemented by a parameterless "self-similarity threshold." Experimental evidence shows that the proposed method also detects occlusions and incoherent motions due to vehicles and pedestrians in non simultaneous stereo.
Anglais
10/2011

IEEE Transactions on Pattern Analysis and Machine Intelligence (IEEE Trans Pattern Anal Mach Intell)
Publisher Institute of Electrical and Electronics Engineers (IEEE)
ISSN 0162-8828 
internationale
05/2012
34
5
930-942

Stereo – block-matching – Number of False Alarms (NFA) – a contrario detection
I.4.8

CNES, ECOS Sud U06E01, STIC Amsud (11STIC-01 - MMVPSCV), InCo - Facultad de Ingenieria - UdelaR.
11286
Liste des fichiers attachés à ce document : 
PDF
Sabater_final_tpami.pdf(828 KB)
TEX
biblio_article.bib(17.3 KB)
Sabater_final_tpami.tex(76.4 KB)
Images
Figure10_1.eps(306.9 KB)
Figure10_2.eps(3.3 KB)
Figure10_3.eps(3.8 KB)
Figure10_4.eps(306.9 KB)
Figure10_5.eps(86.3 KB)
Figure10_6.eps(4 KB)
Figure11_1.eps(109 KB)
Figure11_2.eps(323.5 KB)
Figure11_3.eps(108.9 KB)
Figure11_4.eps(323.5 KB)
Figure1_1.eps(135.4 KB)
Figure1_10.eps(27.4 KB)
Figure1_2.eps(27.4 KB)
Figure1_3.eps(27.4 KB)
Figure1_4.eps(27.4 KB)
Figure1_5.eps(27.4 KB)
Figure1_6.eps(27.4 KB)
Figure1_7.eps(27.4 KB)
Figure1_8.eps(27.4 KB)
Figure1_9.eps(27.4 KB)
Figure2a_1.eps(27.4 KB)
Figure2a_10.eps(27.4 KB)
Figure2a_11.eps(27.4 KB)
Figure2a_12.eps(27.4 KB)
Figure2a_13.eps(27.4 KB)
Figure2a_14.eps(27.4 KB)
Figure2a_15.eps(27.4 KB)
Figure2a_16.eps(27.4 KB)
Figure2a_2.eps(27.4 KB)
Figure2a_3.eps(27.4 KB)
Figure2a_4.eps(27.4 KB)
Figure2a_5.eps(27.4 KB)
Figure2a_6.eps(27.4 KB)
Figure2a_7.eps(27.4 KB)
Figure2a_8.eps(27.4 KB)
Figure2a_9.eps(27.4 KB)
Figure2b_1.eps(27.4 KB)
Figure2b_10.eps(27.4 KB)
Figure2b_11.eps(27.4 KB)
Figure2b_12.eps(27.4 KB)
Figure2b_13.eps(27.4 KB)
Figure2b_14.eps(27.4 KB)
Figure2b_15.eps(27.4 KB)
Figure2b_16.eps(27.4 KB)
Figure2b_2.eps(27.4 KB)
Figure2b_3.eps(27.4 KB)
Figure2b_4.eps(27.4 KB)
Figure2b_5.eps(27.4 KB)
Figure2b_6.eps(27.4 KB)
Figure2b_7.eps(27.4 KB)
Figure2b_8.eps(27.4 KB)
Figure2b_9.eps(27.4 KB)
Figure3.eps(18 KB)
Figure4_1.eps(8.4 KB)
Figure4_2.eps(8.7 KB)
Figure4_3.eps(9.2 KB)
Figure4_4.eps(9.2 KB)
Figure4_5.eps(9.2 KB)
Figure4_6.eps(9.2 KB)
Figure5.eps(21.6 KB)
Figure6_1.eps(83.9 KB)
Figure6_2.eps(246.1 KB)
Figure6_3.eps(242.8 KB)
Figure7_1.eps(95.7 KB)
Figure7_2.eps(276 KB)
Figure7_3.eps(95.7 KB)
Figure7_4.eps(276 KB)
Figure8_1.eps(135.4 KB)
Figure8_2.eps(395 KB)
Figure8_3.eps(26.5 KB)
Figure9_1.eps(327.1 KB)
Figure9_10.eps(3.6 KB)
Figure9_2.eps(327.1 KB)
Figure9_3.eps(3.3 KB)
Figure9_4.eps(32.8 KB)
Figure9_5.eps(71.8 KB)
Figure9_6.eps(71.8 KB)
Figure9_7.eps(209.5 KB)
Figure9_8.eps(209.6 KB)
Figure9_9.eps(209.6 KB)
PS
Sabater_final_tpami.ps(6.4 MB)