We present a method for estimating surface height directly from a single polarisation image simply by solving a large, sparse system of linear equations. To do so, we show how to express polarisation constraints as equations that are linear in the unknown depth. The ambiguity in the surface normal azimuth angle is resolved globally when the optimal surface height is reconstructed. Our method is applicable to objects with uniform albedo exhibiting diffuse and specular reflectance. We extend it to an uncalibrated scenario by demonstrating that the illumination (point source or first/second order spherical harmonics) can be estimated from the polarisation image, up to a binary convex/concave ambiguity. We believe that our method is the first monocular, passive shape-from-x technique that enables well-posed depth estimation with only a single, uncalibrated illumination condition. We present results on glossy objects, including in uncontrolled, outdoor illumination.

Linear depth estimation from an uncalibrated, monocular polarisation image / Smith, William A. P; Ramamoorthi, Ravi; Tozza, Silvia. - 9912:(2016), pp. 109-125. (Intervento presentato al convegno 14th European Conference on Computer Vision tenutosi a Amsterdam, The Netherlands nel October 11–14, 2016) [10.1007/978-3-319-46484-8_7].

Linear depth estimation from an uncalibrated, monocular polarisation image

TOZZA, SILVIA
2016

Abstract

We present a method for estimating surface height directly from a single polarisation image simply by solving a large, sparse system of linear equations. To do so, we show how to express polarisation constraints as equations that are linear in the unknown depth. The ambiguity in the surface normal azimuth angle is resolved globally when the optimal surface height is reconstructed. Our method is applicable to objects with uniform albedo exhibiting diffuse and specular reflectance. We extend it to an uncalibrated scenario by demonstrating that the illumination (point source or first/second order spherical harmonics) can be estimated from the polarisation image, up to a binary convex/concave ambiguity. We believe that our method is the first monocular, passive shape-from-x technique that enables well-posed depth estimation with only a single, uncalibrated illumination condition. We present results on glossy objects, including in uncontrolled, outdoor illumination.
2016
978-3-319-46483-1
978-3-319-46484-8
Linear depth estimation from an uncalibrated, monocular polarisation image / Smith, William A. P; Ramamoorthi, Ravi; Tozza, Silvia. - 9912:(2016), pp. 109-125. (Intervento presentato al convegno 14th European Conference on Computer Vision tenutosi a Amsterdam, The Netherlands nel October 11–14, 2016) [10.1007/978-3-319-46484-8_7].
File in questo prodotto:
File Dimensione Formato  
Tozza_Linear-depth-estimation-frontespizio_2016.pdf

non disponibili

Descrizione: Front matter ottenuto dal link indicato nel campo "Note", che include l'indice del volume.
Tipologia: Altro materiale allegato
Licenza: Accesso privato/ristretto
Dimensione 176.15 kB
Formato Adobe PDF
176.15 kB Adobe PDF   Visualizza/Apri   Richiedi una copia
Tozza_Linear-depth-estimation-Copertina_2016.pdf

non disponibili

Descrizione: Copertina del volume
Tipologia: Altro materiale allegato
Licenza: Accesso privato/ristretto
Dimensione 306.84 kB
Formato Adobe PDF
306.84 kB Adobe PDF   Visualizza/Apri   Richiedi una copia
Tozza_Linear-depth-estimation-BackMatter_2016.pdf

non disponibili

Descrizione: Back Matter scaricato dal link indicato nel campo "Note" e che conclude il volume
Tipologia: Altro materiale allegato
Licenza: Accesso privato/ristretto
Dimensione 101.63 kB
Formato Adobe PDF
101.63 kB Adobe PDF   Visualizza/Apri   Richiedi una copia
Tozza_Linear-depth-estimation-Supplemento_2016.pdf

non disponibili

Descrizione: Materiale supplementare, non presente nel volume cartaceo ma disponibile solo online. Il download del pdf è possibile dal link indicato nel campo "Note" solo per utenti autorizzati.
Tipologia: Documento in Post-print
Licenza: Accesso privato/ristretto
Dimensione 449.23 kB
Formato Adobe PDF
449.23 kB Adobe PDF   Visualizza/Apri   Richiedi una copia
Tozza_linear-depth-estimation_2016.pdf

non disponibili

Descrizione: Articolo principale
Tipologia: Documento in Post-print
Licenza: Accesso privato/ristretto
Dimensione 5.95 MB
Formato Adobe PDF
5.95 MB Adobe PDF   Visualizza/Apri   Richiedi una copia

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11588/759565
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 57
  • ???jsp.display-item.citation.isi??? 40
social impact