A python script to stack BVRI image data into a colorized image. This is adapted from code originally written by Yukei Murakami in the Repo: fits_quickRGB. This code is just a semi-automated script version of that code.
Start by placing your .fits files in the same directory as this script. You should have 1 B filter image, 1 V filter image, and 1 R filter image in the same directory formatted such that the directory looks something like:
n220323.d138.2022crv.R.fit
n220323.d137.2022crv.V.fit
n220323.d136.2022crv.B.fit
The I filter is not needed for this script. We also note that the naming format should be *.obj_name.filter.fit Once the data is properly loaded into the same directory as the script. Just run the following on the command line:
python3 colorizer.py -obj_name 2022crv
Note the name given must match the name in the files.
By default the code performs saturation clipping, to avoid this add a command line argument -no_sat_clip and set it to True. You can also adjust the saturation of the image by including the -sat_factor variable and setting it to an integer between 1 and 10. 10 will give you the most saturated/exposed image, while 1 will give you the dimmest, 5 is the default.
python3 colorizer.py -obj_name 2022crv -sat_factor 5 -no_sat_clip True
One can also add arguments -noise_level and -bg_level to adjust the lower limit of the noise threshold and the upper limit of the background color threshold. The default noise_level is set to 0.8 and the default bg_level is set to 0.3. To add these type the following in the command line:
python3 colorizer.py -obj_name 2022crv -sat_factor 5 -no_sat_clip True -noise_level 0.8 -bg_level 0.3
An image should appear in your directory as a saved png file with the name obj_name.png. By default the code also saves the image data in the form of a numpy array with the name obj_name_img_array.npy which can be loaded in with a simple np.load() call and plotted using plt.imshow(). This can be used for more sophisticated plotting and markups. Below is an example of what the initial input starts as and what the final result is.
The sample data is provided by the Filippenko Group at UC Berkeley originally taken by myself and Kingsley Ehrich on the Nickel 1m Telescope at Lick Observatory.
