[PYTHON] Make an RGB 3 color composite diagram

What is this

I will show you how to make a three-color composite diagram, which is often used in astronomical, using python.

Module to use

• PIL
• astropy --astroquery (necessary if you want to download image data with python)

table of contents

1. Prepare three types of images: Red, Green, and Blue.
2. Use PIL.Image to synthesize 3 colors
3. Interactively create a three-color composite diagram using rgbimage.py

1. Image preparation

Astronomy manages images in a file format called FITS. You can also download it at SkyView Virtual Observatory. You can also download it from a web browser, but this time I will try to download it automatically using python.

code

download.py

from astroquery.skyview import SkyView
import astropy.units

target = 'M17'
surveys = ['2MASS-J', '2MASS-H', '2MASS-K']
radius = 0.3 * astropy.units.deg
pixels = 1000

hdu = SkyView.get_images(target, surveys, radius=radius, pixels=pixels)
hdu[0].writeto('M17-2MASS-J.fits', clobber=True)
hdu[1].writeto('M17-2MASS-H.fits', clobber=True)
hdu[2].writeto('M17-2MASS-K.fits', clobber=True)

A little introduction

M17 is Omega It is a diffused nebula known as a nebula, which is a region where NGC6618 clusters containing many massive stars ionize the surrounding molecular gas. It is a fairly active massive star-forming region in the Milky Way galaxy.

2MASS is a near-infrared all-sky survey observation project. Data for the J band (1.2 μm), H band (1.7 μm), and Ks band (2.2 μm) are available.

2. Make a three-color composite diagram using PIL.Image

Create a three-color composite diagram using the three downloaded data. As a procedure,

1. Read the FITS file
2. Determine the range of min and max and convert to 0-255 int type
3. Save the image I will continue to do.

code

plot.py

import numpy
import PIL.Image
import astropy.io.fits

r = astropy.io.fits.open('M17-2MASS-K.fits')[0].data[::-1,:]
g = astropy.io.fits.open('M17-2MASS-H.fits')[0].data[::-1,:]
b = astropy.io.fits.open('M17-2MASS-J.fits')[0].data[::-1,:]

pix_y, pix_x = r.shape


def scale(data, scale_min=None, scale_max=None, stretch='linear'):
    if stretch == 'linear':
        scale_func = scale_linear
    elif stretch == 'log':
        scale_func = scale_log
    else:
        scale_func = scale_linear
        pass

    if scale_min is None:
        scale_min = numpy.nanmin(data)
        pass

    if scale_max is None:
        scale_max = numpy.nanmax(data)
        pass

    scaled = scale_func(data, scale_min, scale_max)
    scaled = numpy.uint8(scaled * 255)
    return scaled

def scale_linear(data, scale_min, scale_max):
    print('%f, %f'%(scale_min, scale_max))
    scaled = (data - scale_min) / (scale_max - scale_min)
    scaled[numpy.where(scaled<0)] = 0
    scaled[numpy.where(scaled>=1)] = 1
    return scaled

def scale_log(data, scale_min, scale_max):
    print('%e, %e'%(scale_min, scale_max))
    scaled = scale_linear(data, scale_min, scale_max)
    scaled = numpy.log10((scaled * 9) + 1)
    return scaled


img = numpy.zeros([pix_y, pix_x, 3], dtype=numpy.uint8)
img[:,:,0] = scale(r, 630, 840, 'log')
img[:,:,1] = scale(g, 595, 730, 'log')
img[:,:,2] = scale(b, 150, 190, 'log')

PIL.Image.fromarray(img).save('rgbimage-M17-2MASS.png')

An image like this will appear.

3. Interactively create a 3-color composite diagram

Specify the range, check the image, specify the range, and so on to find a good color scheme. It's hard to repeat a lot, so I made a tool that can be operated with GUI.

rgbimage.py

How to use

--Download and save rgbimage.py --Run: python rgbimage.py --Load the file, specify the range and save

When you run it, you will see a window like this.

--Click "Open" and a dialog will open where you can select the FITS file --Option menu labeled "linear": Select stretch from'linear','log','sqrt' -"min text box": Lower limit of range specification bar -"max text box": Upper limit of range specification bar -"min slide bar": Specifies the minimum color range -"max slide bar": Specifies the maximum color range

It looks like this when various settings are made.

Finally, click "Save as" to save the image.

4. Add coordinates to the created image

Try plotting with matplotlib using astropy.wcs.

code

plot_with_coordinate.py

import astropy.io.fits
import astropy.wcs
import astropy.units
import matplotlib.pyplot
import PIL.Image

matplotlib.rcParams['font.family'] = 'arial'

hdu = astropy.io.fits.open('M17-2MASS-K.fits')[0]
wcs = astropy.wcs.WCS(hdu.header)
img = PIL.Image.open('rgbimage-M17-2MASS.png')

fig = matplotlib.pyplot.figure()
ax = fig.add_subplot(111, projection=wcs)
ax.imshow(img)
ax.grid(color='w', linestyle='-')
ax.coords[0].set_major_formatter('hh:mm:ss')
ax.coords[1].set_major_formatter('dd:mm')
ax.set_xlabel('R.A. (J2000)')
ax.set_ylabel('Dec. (J2000)')

ax2 = ax.get_coords_overlay('galactic')
ax2[0].set_ticks(spacing=0.05*astropy.units.deg)
ax2[1].set_ticks(spacing=0.05*astropy.units.deg)
ax2[0].set_major_formatter('d.dd')
ax2[1].set_major_formatter('d.dd')
ax2[0].set_axislabel('Galactic Longitude')
ax2[1].set_axislabel('Galactic Latitude')
ax2.grid(color='y', linestyle='-')

fig.savefig('rgbimage-M17-2MASS-withcoord.png')

An image like this appears.