Using spectrometry and photography to study color underwater

Derya Akkaynak, Eric Chan, Justine J. Allen, Roger T. Hanlon

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Most underwater images are post-processed to look pleasing to human viewers. This often results in unrealistically saturated colors. Images taken for the purpose of studying color-sensitive topics such as marine animal coloration, must represent colors as accurately as possible and should not be arbitrarily enhanced. This first requires a transformation of colors from the camera color space to a device independent space. In this paper we present a method for transforming raw camera-RGB colors to a device independent space, optimizing this transformation for a particular underwater habitat. We have conducted an extensive study of the variation of color appearance underwater at a dive site in the Aegean Sea by taking 21 sets of spectrometry and irradiance readings with corresponding photographs of four different color standards. Spectral and photographic data were collected in the presence of natural daylight at various depths and under different weather conditions. In addition to the color charts, we have built a habitat chart to optimize this camera-specific transformation for a given dive site.

Original languageEnglish
Title of host publicationOCEANS'11 - MTS/IEEE Kona, Program Book
PublisherIEEE Computer Society
ISBN (Print)9781457714276
DOIs
StatePublished - 2011
Externally publishedYes
EventMTS/IEEE Kona Conference, OCEANS'11 - Kona, HI, United States
Duration: 19 Sep 201122 Sep 2011

Publication series

NameOCEANS'11 - MTS/IEEE Kona, Program Book

Conference

ConferenceMTS/IEEE Kona Conference, OCEANS'11
Country/TerritoryUnited States
CityKona, HI
Period19/09/1122/09/11

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Ocean Engineering

Fingerprint

Dive into the research topics of 'Using spectrometry and photography to study color underwater'. Together they form a unique fingerprint.

Cite this