Computational color combination analysis of Papilionidae butterflies as aesthetic objects

Erina Kakehashi*, Keiichi Muramatsu, Haruo Hibino

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

4 Citations (Scopus)


In this study, our aim is to clarify the color combination rules of the human-preferred Papilionidae butterflies as aesthetic objects. A set of 118 butterfly images, including color polyphenism from the 47 Papilionidae species that are generally preferred by humans, was selected. These images were classified using hierarchical cluster analysis based on similarities of lightness, chroma, and hue attributes in CIELAB space, determined using histogram intersection. Then, the color distributions and combinations in each cluster were analyzed using a Gaussian mixture model and the color combination types defined in the present study. Accordingly, we obtained the following main color combination rules of human-preferred Papilionidae: (a) dominant low lightness and contrasting lightness components, (b) dominant low chroma and similar chroma components, and (c) dominant orange to yellow-green hue and similar hue components. These rules partly agree with the robust harmony principles found in previous research. We infer that the cognitive effects concerning the processing fluency through these color combination rules influence human aesthetic responses.

Original languageEnglish
Pages (from-to)65-84
Number of pages20
JournalColor Research and Application
Issue number1
Publication statusPublished - 2020 Feb 1
Externally publishedYes


  • Gaussian mixture model
  • Papilionidae butterfly
  • color harmony
  • image classification
  • quantification of color combinations

ASJC Scopus subject areas

  • Human Factors and Ergonomics
  • Chemistry(all)
  • Chemical Engineering(all)


Dive into the research topics of 'Computational color combination analysis of Papilionidae butterflies as aesthetic objects'. Together they form a unique fingerprint.

Cite this