Detecting Filaments in the Cosmic Web: A New Method for Extremely Noisy Data

What is it about?

iPHT (Iterative Probabilistic Hough Transform) is a new method that detects thread-like structures in extremely noisy environments across any type of dataset. We applied iPHT to the problem of detecting filaments in the cosmic web of galaxies, where galaxies are not randomly distributed but form a vast network of filaments, walls, and voids. iPHT tackles this problem using probabilistic voting and entropy-based filtering to detect filaments even under extreme noise conditions. iPHT consistently outperforms existing clustering and filament detection methods. It is robust, interpretable, and governed by a single physically meaningful parameter, making it a reliable tool for scientific structure discovery in astronomy and beyond.

Featured Image

Photo by Sonal Gupta on Unsplash

Photo by Sonal Gupta on Unsplash

Why is it important?

Understanding how galaxies are distributed across the universe is one of the biggest challenges in modern cosmology. Filaments, the thread-like structures connecting galaxy clusters, play a key role in shaping the large-scale structure of the universe, yet detecting them from noisy observational data remains extremely difficult. iPHT addresses this challenge by providing a robust, reliable, and interpretable method that works even under extreme noise conditions. Beyond astronomy, iPHT has the potential to be applied in any scientific field where hidden thread-like structures need to be discovered in noisy and complex datasets.

Perspectives