Astuvansalmi in Ristiina, Finland, has the largest known prehistoric rock painting collection in the Nordic countries. Some 80 motifs of animals, humans, and abstract figures have been discovered on the cliff that faces Lake Yövesi and date back about 5,000 years. Researchers from the University of Eastern Finland are using advanced imaging and artificial intelligence to reveal paintings that may still lie hidden beneath the surface.
The project also integrates archaeology, computer science, and imaging analysis to document the site in unprecedented detail. With the use of hyperspectral cameras, researchers are able to see subtle differences in pigment composition that are not apparent to the naked eye. High-resolution images enable scientists to differentiate between different shades of red ochre and determine whether different mixtures of paint were used by artists over time.
The entire rock face is also being recorded digitally by a panoramic GigaPan photography, a three-camera scanning rig, and a Matterport 3D scanner. Difficult-to-reach places are photographed with hyperspectral and system cameras, while drone imaging is used when weather permits. All the data collected are uploaded to a cloud server to create an accurate digital twin of the paintings. This digital copy makes it possible to analyze and preserve the fragile art, which has endured harsh environmental conditions over the years.
The ancient landscape is also being reconstructed by researchers using gaming technology. The water level in Lake Saimaa was higher thousands of years ago, and prehistoric artists likely painted while standing on ice or in boats. The scientists virtualize the water level to its former state, thereby visualizing the environment as it was and identifying new locations where undiscovered rock paintings might exist.
As the pigments contain no datable organic material, researchers have replicated experimental paint mixtures—ochre mixed with fat, egg white, or blood—to learn about how ancient emulsions behaved and aged. The experiments enable prehistoric paint composition and application techniques to be interpreted without damaging the original artwork.
Artificial intelligence is increasingly being employed to detect faint motifs and recurring forms masked by erosion. Neural networks will be able to detect shapes like those at other northern rock art sites, such as Alta or Murmansk. This may help trace stylistic links between regions, but scientists are being cautious: AI models trained on unrelated datasets could misinterpret them as cultural or chronological differences.
In addition to discovery, digital preservation makes the art of Astuvansalmi available to the public. Virtual reality applications now allow museum visitors to explore 3D reconstructions of rock painting environments, viewing the paintings as they would have been seen millennia ago.
More information: University of Eastern Finland
