Black edges in registered images after WBPP

Hello everyone,

After executing WBPP script, only for registration purposes, there's some registered images that show slim dark edges on two or three sides of the image. Besides of this, all the images taken after the meridian flip show a big black border in two sides of the image. I've attached three images: one before meridian flip that is ok, another before MF that shows the slim black edges (right and bottom), and the last one is an example of an after-meridian flip image with big black borders. All my images have been calibrated, cosmetic corrected and registered with de WBPP script. I tried to correct the bad images with the StarAlignment process, with a Thin Plate Splines model and choosing the Distortion Correction. The original 58 subs don't show any particular thing that leads me to think that there's any problem with them. I didn't change the orientation of my camera after the meridian flip, so I assume that my flats should be ok for both sides of the meridian. I don't want to integrate my images until this problem is solved, so may someone please help me to solve this situation?

Thank you very much for you help and advise,

Jordi

Comments

  • Hi Jordi,

    This is an outcome of aligning (registering using StarAlignment) your images. One image is your reference frame and all other images are aligned to it. Your images are shifted (due to dithering, pointing or other things) so the regions where they do not overlap become black pixels. 

    This is important since these black pixels, by default, will not be included in the final integrated average of your images (they are rejected). If the normalization is good between your frames and you have enough of them- you will not see the black in your integrated final image.  demonstrate this in a number of workflow videos.

    To understand how rejection works- and more specifically how a pixel stacks can have different numbers of values to average together- please see my videos on the Image Integration Primer and Image Integration.
    It is a lot to absorb...but fundamental to answering a question exactly like this.

    -the Blockhead
  • I should add, if your images *were* very closely aligned..and then after registration became mis-aligned...this is truly a registration issue. Sometimes the distortion correction finds a solution that isn't valid.
    -the Blockhead
  • Mr. Block,

    Thank you very much for your help and advise. I'll give it a try and see if I can get rid of this situation. I understand that one can try to solve this problem aplying a Normalization process, maybe the NSG script, isn't it?
    I will review the Image Integration videos that you said and try to identify the problem.

    Thank you very much,

    Jordi
  • Jordi,

    This isn't a problem though, it is normal when registering images. 
    It IS a problem if the images are not being properly registered.

    -the Blockhead

  • Mr Block,

    Thank you for your explanation and patience. I firstly misunderstood some of your words. I see now that there's no problem, always the images are finally well registered. That can be acomplished if the normalization process is good enough and if I have enough frames to "compensate" the black borders. Is it true? In fact, I applied NSG and then integrated the normalized frames and the final image is ok, so thank you again for your help.

    Jordi
  • Yes.. you have it now... everything is good.
    -the Blockhead
Sign In or Register to comment.