Please use this identifier to cite or link to this item: http://hdl.handle.net/10553/35433
Title: Automatic correction of perspective and optical distortions
Authors: Santana-Cedrés, Daniel 
Gomez, Luis 
Alemán-Flores, Miguel 
Salgado de la Nuez, Agustín Javier 
Esclarín, Julio 
Mazorra, Luis 
Alvarez, Luis 
UNESCO Clasification: 220990 Tratamiento digital. Imágenes
120601 Construcción de algoritmos
120602 Ecuaciones diferenciales
120326 Simulación
Keywords: Lens distortion
Vanishing points
Perspective correction
Issue Date: 2017
Journal: Computer Vision and Image Understanding 
Abstract: Perspective and optical (lens) distortions are aberrations of very different nature that can simultaneously affect an image. Perspective distortion is caused by the position of the camera, especially when it is too close to the scene. Optical distortion is a lens aberration which causes straight lines in the scene to be projected onto the image as distorted lines. Standard methods to correct perspective distortion are based on the estimation of the vanishing points, which can fail if lens distortion is significant. In this paper, we introduce a new method which addresses both problems in a single framework. First we estimate a lens distortion model by extracting a collection of distorted lines in the image. These distorted lines are afterward rectified by means of the lens distortion model and used to estimate the vanishing points. Finally, the vanishing points are used to correct the perspective distortion. We present a variety of experiments to show the reliability of the proposed method.
URI: http://hdl.handle.net/10553/35433
ISSN: 1077-3142
DOI: 10.1016/j.cviu.2017.05.016
Source: Computer Vision and Image Understanding[ISSN 1077-3142],v. 161, p. 1-10
Appears in Collections:Artículos
Show full item record

SCOPUSTM   
Citations

14
checked on Jun 13, 2021

WEB OF SCIENCETM
Citations

10
checked on Jun 6, 2021

Page view(s)

59
checked on Jun 13, 2021

Google ScholarTM

Check

Altmetric


Share



Export metadata



Items in accedaCRIS are protected by copyright, with all rights reserved, unless otherwise indicated.