Perception of prominent colors in natural images : research data underlying the article

Weingerl, Primož (author)

Natural images contain a wide variety of colors, yet little is known about which colors observers perceive as visually prominent and how much they agree on them. This dataset contains data collected to investigate the perception of visually prominent colors in natural images. Using a direct selection paradigm, observers evaluated 56 diverse images under controlled, standardized conditions and identified the colors they perceived as most visually prominent. The dataset includes the images used in the experiments, observer data and color selections for both experiments (E1 and E2), observer characteristics, and candidate colors for each image together with their properties and corresponding numerical values.

• Language: English
• Keywords: color perception, natural images, dominant colors, color salience, color extraction, visual attention
• Typology: 2.20 - Complete scientific database of research data
• Organization: NTF - Faculty of Natural Sciences and Engineering
• Year: 2026
• PID: 20.500.12556/RUL-186069
• Data col. methods: Measurements and tests
• Publication date in RUL: 26.08.2026

Licences

• License: ODC-By, Open Data Commons Attribution License
• Link: https://opendatacommons.org/licenses/by/index.html
• Description: You are free to share, create and adapt work with this license as long as you attribute public use of the database, or works produced from the database.

Secondary language

• Language: Slovenian
• Title: Zaznavanje prominentnih barv naravnih slik : raziskovalni podatki, obravnavani v članku

Abstract

Naravne slike vsebujejo veliko različnih barv, vendar je malo znanega o tem, katere barve opazovalci zaznavajo kot vizualno prominentne (izstopajoče) in v kolikšni meri se o tem strinjajo. Raziskava podaja celovito analizo zaznave vizualno prominentnih barv naravnih slik, pri čemer se osredotoča na njihovo število, ujemanje med opazovalci, raznolikost ter povezavo s posameznimi barvnimi atributi (kroma, svetlost in barvni ton). Opazovalci so pod kontroliranimi, standardiziranimi pogoji iz nabora 56 raznolikih slik izbrali barve, ki so jih zaznali kot najbolj vidne oziroma izstopajoče. Nabor raziskovalnih podatkov vključuje slike, uporabljene v eksperimentih, podatke o opazovalcih in njihovih izbirah barv za oba eksperimenta (E1 in E2), lastnosti opazovalcev ter za vsako sliko nabor barv kandidatk skupaj z njihovimi lastnostmi in pripadajočimi numeričnimi vrednostmi.

• Keywords: zaznavanje barv, naravne slike, dominantne barve, barvna salientnost, ekstrakcija barv, vizualna pozornost

Projects

• Funder: ARIS - Slovenian Research and Innovation Agency
• Project number: P2-0450
• Name: Inovativna grafična tehnologija