Recovering Depth from Still Images for Underwater Dehazing Using Deep Learning
Ver/ Abrir
Impacto
Scholar |
Otros documentos de la autoría: Pérez Soler, Javier; Bryson, Mitch; Williams, Sephan B.; Sanz, Pedro J
Metadatos
Mostrar el registro completo del ítemcomunitat-uji-handle:10234/9
comunitat-uji-handle2:10234/7036
comunitat-uji-handle3:10234/8620
comunitat-uji-handle4:
INVESTIGACIONMetadatos
Título
Recovering Depth from Still Images for Underwater Dehazing Using Deep LearningFecha de publicación
2020Editor
MDPIISSN
1424-8220Cita bibliográfica
PÉREZ, Javier, et al. Recovering Depth from Still Images for Underwater Dehazing Using Deep Learning. Sensors, 2020, vol. 20, núm. 16, p. 4580Tipo de documento
info:eu-repo/semantics/articleVersión de la editorial
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7472610/Versión
info:eu-repo/semantics/publishedVersionPalabras clave / Materias
Resumen
Estimating depth from a single image is a challenging problem, but it is also interestingdue to the large amount of applications, such as underwater image dehazing. In this paper, a newperspective is provided; by ... [+]
Estimating depth from a single image is a challenging problem, but it is also interestingdue to the large amount of applications, such as underwater image dehazing. In this paper, a newperspective is provided; by taking advantage of the underwater haze that may provide a strong cue tothe depth of the scene, a neural network can be used to estimate it. Using this approach the depthmapcan be used in a dehazing method to enhance the image and recover original colors, offering abetter input to image recognition algorithms and, thus, improving the robot performance duringvision-based tasks such as object detection and characterization of the seafloor. Experiments areconducted on different datasets that cover a wide variety of textures and conditions, while using adense stereo depthmap as ground truth for training, validation and testing. The results show that theneural network outperforms other alternatives, such as the dark channel prior methods and it is ableto accurately estimate depth from a single image after a training stage with depth information. [-]
Publicado en
Sensors, 2020, vol. 20, núm. 16, p. 4580Proyecto de investigación
This work has been partly funded by Spanish Ministry under DPI2014-57746-C3 grant (MERBOTS Project) and DPI2017-86372-C3 grant (TWINBOT Project); by Valencian Government under IDIFEDER/2018/013 grant (CIRTESU Project); and by Universitat Jaume I under PREDOC/2012/47 grant and UJI-B2018-34 grant (NEPTUNO Project).Derechos de acceso
info:eu-repo/semantics/openAccess
Aparece en las colecciones
- ICC_Articles [418]
El ítem tiene asociados los siguientes ficheros de licencia:
Excepto si se señala otra cosa, la licencia del ítem se describe como: © 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open accessarticle distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).