Segmentation problem (in comparison to sExtractor)

[igorkolesnikov13]

While conduction some tests, I noticed very strange behavior on the segmented map. The summary is that SEP, in some cases, fragments the galaxy in several segments, while sExtractor is able to maintain the galaxy in one piece. Example annexed.

I know that I can decrease the threshold to improve it, but it is not always possible, because I am trying to find generic value for several thousands of galaxies.

Is there some possibility to force SEP to maintain galaxy integrity?