Color Name Applications in Computer Vision
Parraga, Carlos Alejandro ![ORCID Identifier](/img/uab/orcid.ico)
(Universitat Autònoma de Barcelona. Departament de Ciències de la Computació)
Akbarinia, Arash ![ORCID Identifier](/img/uab/orcid.ico)
(Justus-Liebig-Universität Gießen)
Shamey, Renzo ed.
Imprint: |
Berlin, Heidelberg : Springer, 2020 |
Description: |
7 pàg. |
Abstract: |
In Computer Vision, the association of names to colors is one of the fundamental problems in the field of image understanding. There are numerous computational applications (e. g. image retrieval, visual tracking, person identification, human-machine interaction, etc. ) that require pixels to be labelled according to the color perceived by the user. This is relatively easy for focal colors under canonical illuminants, where the agreement is high, but becomes increasingly difficult as perceptions move away from these conditions. For these difficult cases, the traditional solution tends to be a collection of "ad-hoc" strategies, however, new approaches that combine knowledge from anthropology, linguistics, visual perception and machine learning have offered promising results. Specifically, deep neural networks appear to possess all the required building blocks to offer a color naming solution "in the wild". This article reviews the current state of knowledge and discusses open challenges with a multidisciplinary (and non-specialized) readership in mind. |
Rights: |
Tots els drets reservats. ![](/img/licenses/InC.ico) |
Language: |
Anglès |
Document: |
Capítol de llibre ; recerca ; Versió acceptada per publicar |
Published in: |
Encyclopedia of Color Science and Technology, 2020, p. 1-7, ISBN 9783642278518 |
DOI: 10.1007/978-3-642-27851-8_404-1
The record appears in these collections:
Books and collections >
Book chapters
Record created 2023-05-23, last modified 2024-06-14