Skip to main navigation Skip to search Skip to main content

Short communication: White balance for self-luminous highlights in images

Research output: Journal article publicationJournal articleAcademic researchpeer-review

Abstract

An image is commonly white balanced using a white point that is considered to accurately characterize the scene illumination color (e.g., the white point estimated by the main camera on a smartphone, or the white point estimated by an ambient light sensor). Two recent studies, however, suggest that the perceived color appearance of self-luminous highlights is less affected by the ambient illumination color in a scene, with the color appearance judged using a white point between 5000 and 7000 K, especially when the stimulus luminance is much higher than the diffuse white luminance. Based on this, we hypothesized to use a D65 white point for white balancing the self-luminous highlights in images, regardless of the estimated white points. We captured a series of images containing self-luminous highlights, and processed them with three different white points (i.e., the white point estimated by the main camera, D50, and D65). The hypothesis was supported by the survey results collected from 73 observers, together with our observations, and believed to introduce better image color appearance, especially for images shown on high dynamic range displays.

Original languageEnglish
Pages (from-to)119-124
Number of pages6
JournalColor Research and Application
Volume50
Issue number2
DOIs
Publication statusPublished - 1 Mar 2025

Keywords

  • chromatic adaptation
  • white balance

ASJC Scopus subject areas

  • Human Factors and Ergonomics
  • General Chemistry
  • General Chemical Engineering

Fingerprint

Dive into the research topics of 'Short communication: White balance for self-luminous highlights in images'. Together they form a unique fingerprint.

Cite this