AI Model Identified Pigments in Historical Paintings
Researchers developed a machine learning tool that classifies pigments in art with 99.3 percent accuracy.
Updated on Oct. 5, 2026 in Artificial Intelligence
Live Poll
Do you trust artificial intelligence to accurately identify and help preserve historical artifacts?
In 2025, researchers introduced a new machine learning framework capable of identifying pigments in historical paintings. By leveraging a random forest model and multi-spectral imaging, the tool achieved a classification accuracy of 99.3 percent.
Why it matters
Traditional methods for analyzing historical pigments often require complex, time-consuming instrumentation or are limited by spatial constraints. This new automated approach streamlines the process, providing a more efficient way to examine art history without heavy data processing.
The classification model functions using three spectral illumination conditions including visible-light, ultraviolet false-colour, and infrared false-colour imaging. It was trained on 4,000 image patches, each measuring 32 x 32 pixels, derived from 600 raw images.
The players
Applied Sciences
This is a peer-reviewed, open-access journal that focuses on applied natural sciences and engineering research.
The details
The classification framework identifies 40 unique pigments across five variants by analyzing feature vectors generated from RGB channels. Researchers validated the model by applying it to annotated regions of historical paintings to ensure practical reliability.
Timeline
The study documenting the model was published in Applied Sciences in 2025.
The Tech Race
This development follows a pattern set by the National Gallery of Art's technical art history research programs by automating non-invasive pigment analysis. It represents a shift from legacy, instrument-heavy hyperspectral imaging toward lighter, software-defined classification tools.
For museum conservators and art historians, this technology offers a faster, more accessible method to authenticate works and study artist techniques without risking damage. It reduces the barrier to entry for detailed chemical analysis of cultural heritage items.
The takeaway
Automated pigment analysis demonstrates how machine learning can bridge gaps in technical fields where manual inspection is traditionally prohibitive. Implementing these algorithms allows for larger-scale preservation and deeper understanding of material history in art.
Further reading
Discover more about the evolving landscape of Artificial Intelligence in current research applications.
Source note: This article includes information reported by European Coatings.
Live Poll
Do you trust artificial intelligence to accurately identify and help preserve historical artifacts?
