A MST algorithm for source detection in γ-ray images

Abstract

The Minimal Spanning Tree (MST) algorithm is a graph-theoretical method of finding clusters in a given set of points. We apply MST to gamma-ray bidimensional images and associate possible sources with the regions where photons' arrival directions clusterize. Additional filters to select clusters are included to reduce spurious detections. We present some applications to simulated GLAST data. © 2007 American Institute of Physics.