Page 1 of 1

image deduction

Posted: Mon Mar 09, 2020 11:28 pm
by kora
Supernova research, especially at the center of galaxies, is a good way to fit and subtract a reference image.
Description: https://authors.library.caltech.edu/706 ... 1006v1.pdf

I first installed VaST Transient search for Linux, but it doesn't know image extraction, it works with photometric lines. A friend of mine, Robert Fridrich, gave him many variables in the SX database, but he was not a supernova patrolman.
http://scan.sai.msu.ru/vast/


Second, I found the FITSH collection of András Pál. Great value! Linux shell and python must write their own script.. It is successfully used in HAT-net and HAT-pi project.
https://fitsh.net/wiki/Example:Main_Page

For galaxies that do not have their own reference image, download a DSS image according to the ATIK 383L + field of view. Simple command line... download image.
dss_ref_download.png
Unfortunately, the Prism version only recognizes a fixed 15'x15' size. Can Be Improved?
dss_image.png
Linux system the local astrometry.net, and sextractor and hotpants program package can extract the image together. But you have to build your own cable on Linux ...
hotpants.png

:?:
QUESTION:
How can you do something similar with the Prism, image matching and subtraction? Prism script writing? I look forward to your advice!

Re: image deduction

Posted: Wed Mar 11, 2020 12:01 am
by kora
First step: In the Prism, I open the reference image and put a test image. The images are centered on the atmosphere and optical distortion and other inaccuracies are not the same.
ref1_and_2photo.png
Second: Second: In the Processing / Image to Image Math menu, I select Substraction.
substarction.png
Third: I select the image in the new window.

Step Four: The result shows the subtraction. Unfortunately, the star profile is visible everywhere.
substarction_result.png
QUESTION: How do I deduce the mismatch of different geometry images? :?:

Hope somebody here has a solution to the good subtraction problem ..


Regards Alexander