본문 바로가기
  • Home

Painterly Rendering with Image Relighting-Based Faded Color

  • Journal of The Korea Society of Computer and Information
  • Abbr : JKSCI
  • 2025, 30(4), pp.55~63
  • Publisher : The Korean Society Of Computer And Information
  • Research Area : Engineering > Computer Science
  • Received : February 25, 2025
  • Accepted : April 17, 2025
  • Published : April 30, 2025

Jong-Hyun Kim 1

1인하대학교

Accredited

ABSTRACT

This paper proposes a painterly rendering method that reflects 2D image relighting and color variation to visualize the effects of light on oil paintings. The proposed method relights a 2D image and uses the resulting shading values as weights to reflect corresponding changes in color. For the relighting process, the 2D image is approximated as a 3D surface, and normals are estimated. The angle between these normals and the specified light source is used to compute shading values. Users can freely set the position of the light source, and the resulting color changes under varying lighting conditions are derived with reference to prior research. Unlike conventional local-based painterly rendering methods that produce flat and overly simplified results, the proposed approach effectively captures faded colors and a sense of depth through image relighting. This allows for vivid, three-dimensional representations that closely resemble real oil paintings. Ultimately, the method contributes to the expressive depiction, prediction, and restoration of color changes in cultural and artistic works.

Citation status

* References for papers published after 2023 are currently being built.

This paper was written with support from the National Research Foundation of Korea.