Документ взят из кэша поисковой машины. Адрес оригинального документа : http://xmm.vilspa.esa.es/sas/5.4.1/doc/ewavelet/node16.html
Дата изменения: Fri Jan 10 22:11:34 2003
Дата индексирования: Sat Dec 22 10:29:05 2007
Кодировка:

XMM-Newton Science Analysis System


ewavelet (ewavelet-3.3) [xmmsas_20030110_1802-5.4.1]

Meta Index / Home Page

Algorithm

The wavelet detection involves the following steps:

• Read the input image and the exposure map.
• Generate the countrate image from the input image and exposure map.
• Generate the initial background map by scaling the exposure map in such a way that the background map contains as many ``counts'' as the input image.
• Do the following for several wavelet scales:

  1

  generate a background map for this particular scale

  2

  convolve image with the MH wavelet function

  3

  find local maxima in the convolved image

  4

  Check if the local maximum is significant

  5

  If the maximum is significant, add the point to the source list and calculate the number of counts from the wavelet correlation value ($C_{i,j}$)

• Determine if some sources were detected at several wavelets scales. If so, choose the wavelet scale with the highest local correlation value.
• Use the detections of the same source at several wavelet scale to get some information on the source properties.
• Make a new iteration or generate the final source list.

Figure: Block diagram of the wavelet source detection algorithm.