Dog Picking » History » Revision 4
Revision 3 (Arne Moeller, 06/03/2010 10:24 AM) → Revision 4/14 (Arne Moeller, 06/03/2010 10:24 AM)
h1. Dog Picking h2. 1. Test first then submit Since Dog Picker does not use any templates you will have to play around with the different settings Dog picker will pick *white blobs*, so make sure to invert the density if necessary (usually ice images) Test the settings on a characteristic image (simply paste the filename of a typical image in the test settings box). To get an idea which parameters suit your data "mouseover" the boxes. There you can find some general estimates which you can use as starting point. During testing you should optimize one parameter at a time and then move on to the next. It is a good idea to optimize the settings on one image and then test a second image. Don’t worry about the final boxsize or binning this will be determined in the next step: [[stacks]] ! h2. 2. Click on Run Dog Picker to submit the job If you submit the job while you are still collecting data use the option: wait for more images after finishing. h2. 3. Continue with [[stacks]] ______ [[Template Picking|< Template Picking]] | [[Manual Picking|Manual Picking >]] ______