Institutional Repository of Vilnius University
Spalvų analizės sistema interjero vizualizacijoms: spalvų išgavimo ir suderinamumo vertinimo sistema /
Abstract
dc:descriptionThe topic of this thesis is a color analysis system for interior visualizations, focused on color extraction and compatibility assessment. The right choice of colors is one of the most important aspects of interior design, affecting not only the mood and feeling of a room, but also a person’s wellbeing and emotional comfort. Coordinated colors help to create a cozy and pleasant environment where people feel comfortable. Properly assessed color compatibility provides clearer guidelines that help users choose colors that match each other more easily and avoid visually incompatible solutions. The aim of the work is to create a system that allows automatic extraction of color palettes from interior photos using different color quantization methods, and provides the possibility to obtain color harmony and compatibility assessments with recommendations. The tasks of the work include the analysis of color quantization methods, experimental comparison, implementation of algorithms, and the application of color harmony assessment methods. The conducted experimental study confirmed that the model utilizing the Mini–Batch K–Means algorithm performs, on average, 13.5 times faster than other standard methods under maximum load (K = 32). Furthermore, statistical analysis demonstrated that iterative quantization methods ensure superior visual quality; however, to balance high accuracy with result reproducibility, a hybrid model integrating both Mini–Batch K–Means and Mean–Cut algorithms was implemented. The developed prototype provides a scientifically grounded tool for analyzing color combinations for a wide range of users – from interior designers to general users. This facilitates the reduction of subjectivity when making aesthetic decisions. Further development involving the integration of texture recognition algorithms would enable a transition from purely colorbased to comprehensive interior analysis.
Degree
thesis:*- Grantor dc:publisher
- Institutional Repository of Vilnius University
- Year dc:date
- 2026
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Petkevičius, Gustas,
Rights
dc:rights- Statement dc:rights
-
- info:eu-repo/semantics/openAccess
- Language dc:language
- lit
Identifiers
dc:identifier.*- Repository record dc:identifier
- https://repository.vu.lt/VU:ELABAETD277690124&prefLang=en_US