Dinosaur footprints are one of the most common types of fossils. Scientists sometimes find a single isolated footprint, while at other times they come across scattered, irregular traces that resemble dancing leaps in a dinosaur's ballroom. However, determining which specific dinosaur left which footprint remains an extremely difficult task.
Researchers are now developing a methodology that uses artificial intelligence to help identify the dinosaur species responsible for each footprint, based on eight different characteristics of the tracks.
As stated by physicist Gregor Hartmann from the German Helmholtz Center research center in Berlin: "This is important because it provides an objective method for classifying and comparing tracks, which reduces reliance on subjective human interpretation." Hartmann is the lead author of the research published in the scientific journal "Proceedings of the National Academy of Sciences".
The new methodology was developed by analyzing 1974 images and drawings of footprint shapes using algorithms, spanning 150 million years of dinosaur history.
Artificial intelligence was able to distinguish eight features that explain the variation in the shapes of these tracks. These features include the total load and shape reflected from the contact area of the foot with the ground, the position of the load, the spread of the toes, how the toes attach to the foot, the position of the heel, the heel load, the relative sharpness of the toes versus the heel, and the difference in shape between the right and left sides of the track.
