Local color correction for web-pages content harmonization
DOI:
https://doi.org/10.59476/ilpmt2026.v2i2.802Keywords:
color correction, color harmonization, semantic segmentation, CIELAB color space, web interfacesAbstract
Modern web interfaces increasingly rely on dynamic visual content such as images, banners, and user-generated media. Unlike static layouts, these systems must maintain visual consistency despite continuous changes in content. A common issue arises from the mismatch between predefined interface color palettes and the arbitrary colors of embedded images, leading to visual inconsistency and reduced aesthetic quality. Existing color harmonization methods mainly focus on transferring color characteristics between images or improving internal coherence. However, they typically ignore the constraints of interface design systems and do not consider the semantic im-portance of different image regions. As a result, such approaches may distort visually significant elements while attempting to achieve global color consistency. This paper proposes a semantic-aware approach to color harmonization of dynamic visual content. The method adapts image colors to a predefined interface palette while preserving the semantic integrity of key elements. It operates in the CIELAB color space, where color relationships can be represented geometrically on the chromatic plane. The proposed method includes several stages. First, the interface palette is analyzed and a base color is select-ed. Then, dominant colors are extracted from the image, and the image is di-vided into semantically significant and secondary regions. A target harmonic scheme is defined using angular relationships relative to the base color. Color transformations are applied selectively, with stronger adjustments in second-ary regions and constrained modifications in important areas. To ensure natural results, harmonic relationships are treated as tolerance regions using the ΔE metric rather than strict constraints. This allows flexible adaptation while maintaining perceptual consistency. The proposed approach supports semi-automatic color correction of dynamic content, reducing the need for manual adjustments and improving visual coherence in web interfaces.
References
1. Wang S., Liu J., Jiang J., Jiang Y. & Lan J. (2022). Attribute analysis and modeling of color harmony based on multi-color feature extraction in real-life scenes. Front. Psychol. 13:945951. doi: 10.3389/fpsyg.2022.945951
2. Weingerl, P. (2024). Automated Image-Based User Interface Color Theme Generation. Applied Sciences, 14(7), 2850. https://doi.org/10.3390/ app14072850
3. Reinhard E., Ashikhmin M., Gooch B. & Shirley P. (2001). Color Trans-fer between Images. IEEE Computer Graphics and Applications, 21, 34-41. 10.1109/38.946629.
4. Chang H., Fried O., Liu Y., DiVerdi S. & Finkelstein A. (2015). Palette-based photo recoloring. ACM Transactions on Graphics, 34, 4, 139:1 – 139:11. https://dl.acm.org/doi/epdf/10.1145/2766978
5. Cong W., Zhang J., Niu L., Liu L., Ling Z., Li W., & Zhang L. (2019). Do-veNet: Deep Image Harmonization via Domain Verification. 2020 IEEE/ CVF Conference on Computer Vision and Pattern Recognition (CVPR), 8391-8400.
6. Lan X., Li Z., Liu Q., Meng Q., Nie L. & Zhang S. (2024). High-Resolu-tion Image Harmonization with Adaptive-Interval Color Transformation. 13769-13793. 10.52202/079017-0441.
7. Itten J. (1999). The Art of Color: The Subjective Experience and Objec-tive Rationale of Color. Reinhold, 1961. 155 p.
8. Fraser B., Bunting F. & Murphy C. (2004). Real World Color Manage-ment (2nd ed.). Pearson Education. 608 p.
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Nonna Kulishova, Daria Harbuzova

This work is licensed under a Creative Commons Attribution 4.0 International License.