Perception of Prominent Colors in Natural Images

Understanding the perception of prominent colors in natural images can support the development of color extraction algorithms that better reflect human perception, a key requirement for a wide range of applications, from image retrieval to palette generation. Since little attention has been devoted to this topic, the present study provides a systematic analysis of the perception of prominent colors, focusing on their number and diversity, inter-observer agreement, and associations between prominence and color attributes. In two experiments under controlled viewing conditions, a total of 293 observers selected colors from predefined palettes of 25 candidate colors for each of 56 images and ranked them by perceived prominence. When three to seven colors could be selected, observers most commonly selected four, and differences between observers accounted for substantially more variance in the number of selected colors than differences between images (ICC = 0.53 vs. 0.10). Inter-observer agreement was well above chance, particularly at the color-category level, with observers generally agreeing on approximately three color categories, and was strongest for the first-ranked color. The selected colors were highly diverse, usually belonging to different color categories. High-chroma colors were selected more frequently than expected from both the image and the candidate colors and ranked higher, whereas low-chroma and dark colors were selected less frequently and were ranked lower. Orange–red hues and colors of intermediate lightness were selected more frequently than expected from their occurrence in the images, but not beyond their representation in the candidate palettes; however, they were ranked significantly higher than expected from random ordering. Because selection was restricted to predefined palettes, the results should be confirmed with unrestricted color selection. These findings reveal consistent patterns but also show that color prominence is a complex, partly observer-dependent phenomenon, likely shaped by multiple interacting color and image properties.

Authors

Institutions

Publication Details

Journal
Journal of Imaging
Published
2026-11-18
DOI
https://doi.org/10.3390/jimaging12110583
Primary Topic
Color perception and design
Type
article
Field-Weighted Citation Impact
0.00

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Perception of Prominent Colors in Natural Images

Primož Weingerl
1 links
Journal of Imaging
Color perception and design
article

Perception of Prominent Colors in Natural Images

Primož Weingerl
article en

Abstract

Understanding the perception of prominent colors in natural images can support the development of color extraction algorithms that better reflect human perception, a key requirement for a wide range of applications, from image retrieval to palette generation. Since little attention has been devoted to this topic, the present study provides a systematic analysis of the perception of prominent colors, focusing on their number and diversity, inter-observer agreement, and associations between prominence and color attributes. In two experiments under controlled viewing conditions, a total of 293 observers selected colors from predefined palettes of 25 candidate colors for each of 56 images and ranked them by perceived prominence. When three to seven colors could be selected, observers most commonly selected four, and differences between observers accounted for substantially more variance in the number of selected colors than differences between images (ICC = 0.53 vs. 0.10). Inter-observer agreement was well above chance, particularly at the color-category level, with observers generally agreeing on approximately three color categories, and was strongest for the first-ranked color. The selected colors were highly diverse, usually belonging to different color categories. High-chroma colors were selected more frequently than expected from both the image and the candidate colors and ranked higher, whereas low-chroma and dark colors were selected less frequently and were ranked lower. Orange–red hues and colors of intermediate lightness were selected more frequently than expected from their occurrence in the images, but not beyond their representation in the candidate palettes; however, they were ranked significantly higher than expected from random ordering. Because selection was restricted to predefined palettes, the results should be confirmed with unrestricted color selection. These findings reveal consistent patterns but also show that color prominence is a complex, partly observer-dependent phenomenon, likely shaped by multiple interacting color and image properties.

Journal of ImagingVol. 12(11)
University of Ljubljana (SI)
Javna Agencija za Raziskovalno Dejavnost RS
Color perception and design
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon