Feature Extraction from Oriental Painting for Wellness Contents Recommendation Services

Mucheol Kim, Dongwann Kang, Namyeon Lee

Research output: Contribution to journalArticlepeer-review

8 Scopus citations

Abstract

As the interest in health increased, people are more interested in mental health as well as physical health. Predominantly, due to the development of IT technology and digital contents, production of wellness contents through fusion with digital contents is increasing. Although many types of research that pursue wellness through the satisfaction of the visual sense are increasing, they were dealing with the western painting that emphasizes color and saturation. On the other hand, oriental painting is different from western painting in color and composition, and the expression is also very subjective. In addition, due to the material characteristics and composition of oriental painting, it is often used for mental health treatments such as mental health and self-growth. In this paper, we analyze characteristics of materials and composition of the oriental painting and propose the feature extraction method suitable for them. We also suggest the oriental painting recommendation approach that can provide users with customized digital contents to support wellness. In the experiment, feature extraction results are compared and the appropriateness of the recommendation results is evaluated. The results of the proposed approach are expected to be utilized as a personalized digital contents recommendation service for mental health management of people in the future.

Original languageEnglish
Article number8685108
Pages (from-to)59263-59270
Number of pages8
JournalIEEE Access
Volume7
DOIs
StatePublished - 2019

Keywords

  • data analytics
  • feature extraction
  • information retrieval
  • oriental painting
  • Recommender system
  • wellness

Fingerprint

Dive into the research topics of 'Feature Extraction from Oriental Painting for Wellness Contents Recommendation Services'. Together they form a unique fingerprint.

Cite this