IMAGE SEGMENTATION ALGORITHM FOR FILAMENTARY STRUCTURES DETECTION
Sarah Baimukhametova, Dana Alina, Adai Shomanov · Eurasian Journal of Mathematical and Computer Applications · 2021
Large observational surveys of space telescopes in the infra-red and submillimeter ranges revealed intriguing structure of the interstellar medium which implied a boost in the development of filament detection algorithms. Thousands of filaments can be detected in astronomical data sets and detailed studies turn to be challenging. We propose a method that aims at building individual maps fully containing at least one coherent filament, out of larger maps. The method is based on the image processing techniques such as the Rolling Hough Transform, the image component labeling, and the center of mass technique. We apply our method to the Planck satellite data and show that known nearby molecular clouds are well identified with our method.