Detecting X-ray sources with the wavelet transform

Publisher Website
Google Scholar
Abstract
We present here preliminary results of a detection algorithm based on the Wavelet Transform . This technique is an alternative to the standard techniques based on Maximum‐likelihood sliding box methods. It consists of an orthonormal decomposition of a signal into both space and scale through a multiscale filtering process. The position, net counts and morphological parameters of the sources can be automatically estimated, once a source model is assumed, without any knowledge of the local background. The detection itself does not require knowledge of the PSF. Simulations are required to assess the significance of the detection. The method works equally well for both point‐like and extended sources. We have tested this technique on both simulations and ROSAT data. The wavelet analysis of the Lockman field, the deepest ROSAT‐PSPC observation, shows that this algorithm is also well suited for particularly crowded fields. We compare our results with those obtained using other techniques. Our algorithm detects several extended sources (i.e. with characteristic scale significantly greater than the point response function of the telescope/detector system) in this field; two of these have recently been optically confirmed as clusters.

This publication has 0 references indexed in Scilit: