A complete guide to astronomical image processing (1): basic technology of solar system.

In this section, you will learn:

& gt image construction technology

& gt using elevations and curves

& gt color processing

& gt enhance planetary details

In the first part of this tutorial, we will learn how to deal with celestial bodies in the solar system, including the sun, the moon and bright planets. We will see how to use levels and curves to adjust the tone of the target. The function of color scale is similar to that of curve. The color scale is a little simpler and can be adjusted by three sliders. The curve is a little more complicated. These two methods are effective for extracting details from dark areas and suppressing bright areas.

We will also learn how to deal with the color of the earth. There is a simple technique to get the correct color filter. Finally, let's discuss the related technologies to improve the details of the planet. No matter what your current level is, I believe you can learn something you are interested in from this article, which will help you on your way to learning astronomical image processing.

Important software:

Photoshop:Photoshop has always been the best graphic editing software on the market, but it is expensive. If you can't afford it, you can also use the simplified version of Photoshop Elements, which has enough functions. For the citizens of China, you can ignore this statement.

Paintshop Pro: This is also a powerful image editing software. Like PS, it has the advantages of high speed, layering and real-time preview.

GIMP: the abbreviation of "GNU image control program" and a powerful editor based on layers. Different from the above two modes, it is free, can be downloaded from www.gimp.org, supports various operating systems, and has a large number of plug-in libraries.

Methods/steps

Image construction technology-synthesis of color image and gray image

The panchromatic photos of the planet are composed of photos taken by red, green and blue filters with high frame rate and monochrome camera. There is still some work to be done before we get the final color photos.

The camera shoots videos in avi format, which need to be processed by some overlay software, such as Registax or avistack (the usage of registax will be explained in detail in the second part). These softwares will select the best frames from avi videos and then superimpose them to get clearer, brighter and more detailed photos. A red, a green and a blue filter. Of course, these photos are all black and white, and you can use them to create color photos.

Ok, the next step starts here. Suppose you have got the superimposed black and white photos taken by the red, green and blue filters, and make them into color photos step by step according to the following steps.

Step 1: Open the red image (that is, it is actually black and white) and modify the size to 200%, so that alignment is easier.

Step 2: Open the green image, modify the size to 200%, select all, copy and paste it into the new layer above the red image in the previous step. The same is true of the blue image, which is copied on top of the green layer.

Step 3: Hide the blue image layer and gently move the green layer to match the red image below. I can't see how the layers below are aligned. The original text is not quite right, at least change the transparency. ) Display the blue layer and move the blue layer to align with the green layer.

Step 4: After moving, there will be blank around. Use the cutting tool to cut off the surrounding blanks.

Step 5: Create a new blank RGB image with the same size as the image cut in the previous step. Paste the red layer in step 4 into the red channel of the blank image.

Step 6: Repeat the previous step and paste the green layer into the green channel and the blue layer into the blue channel. You're finished.

Use Levels and Curves-Adjust the tonal range of the image.

Levels and Curves tools are used to adjust the intensity of dark parts, highlights and midtones of an image. For best results, we should use 16-bit images instead of 8-bit images. These two tools have similar functions, the color code is easier to learn and the curve is more powerful.

A color scale is a histogram that represents the number of pixels of each brightness value. Only three adjustment sliders are needed to adjust the black, white and gray of the image. The black slider defines a threshold, and all pixels darker than this threshold are regarded as pure black; The white slider also defines a threshold, and all pixels brighter than this threshold are considered pure white. The middle slider redefines the middle value between black and white, and changes the middle tone to fit the newly defined range. The curve tool provides more operations, and a line can define multiple adjustment points.

Sometimes, the simplicity of the color scale is enough to get a perfect picture. But when you want to manually improve the dark area while keeping other areas unchanged, the advantages of the curve come out.

planet

Levels and curves can be used to enhance planetary images, such as enhancing contrast, processing dark parts and suppressing highlights.

Ideally, the histogram of the color scale should have a peak from the leftmost side of the coordinate, then rise smoothly and then fall, like a hill. If this "hill" is concentrated near the black slider, the image is underexposed; If it is too close to the white slider, the image will be overexposed.

The mountain peak on the left represents the black sky background around the planet. If the peak is not very close to the edge, use the black color picker and click the area of the sky background to define the background as black.

Move the white slider until the histogram starts to rise from the right. Note that the central area of the planet is not too bright, and then move the slider in the middle to adjust the overall feeling.

The curve tool provides more adjustment methods, just be careful not to destroy the overall balance of the image. Add adjustment points (1/4, 1/2, 3/4) at the quarter of the curve, and then add adjustment points at 1/8, 7/8, and both the dark part and the bright part can be well adjusted.

moon

Color scale and curve tools can extract subtle dark details or repair the information of overexposed areas.

If there is a black sky background around the moon in your photo, click on the background area with a black eyedropper and tell the software to define the background as black. If there is no background, move the black slider to the place where the histogram starts to rise. The shadow in the middle of the crater should not be completely black, but a dark gray.

The histogram shows the number of pixels (ordinate) of each brightness value (abscissa). Black on the left and white on the right. However, if you want to show the details of the brightest part through the white slider, it is usually not very good. It is recommended not to move the white slider.

Curves are best for dealing with details of shadows and highlights. Add adjustment points at a quarter of the curve, and then add adjustment points at 1/8 and 7/8. Drag the adjustment point to pull out the dark details and press the extra area. After adjustment, fine-tune other points to make the overall light and shade more balanced.

Sun 1: all day, white filter

The white sun is not as bright and dark as the moon, and its histogram is smoother. The sun is partly bright and the edges are dark (but not black), so it is best to set the edges to black.

The color coding tool is very useful for adjusting the image of the entire sun surface. First, click the sky background on the edge of the sun with the black color picker, so that the black slider will be set automatically. Drag the white slider to the right of the histogram. Then adjust the middle slider to make the center of the sun's surface brighter without losing details.

The curve tool can better adjust the image, but it is also easier to destroy the delicate balance between light and shade of the sun. The easiest way is to add an adjustment point at a quarter of the curve. Pull the curve out of a slight S-shape, and be careful not to make the edge of the sun too dark and the center of the sun too bright. After finishing, fine-tune the adjustment point in the middle and adjust the overall balance.

If properly adjusted, this slight S-shaped curve can effectively enhance the details of the sun's edge, such as flares. Moreover, the center of the sun's surface is the brightest, and the surrounding area gradually darkens, making the sun look very 3D.

Sun 2: the edge of the sun, Hα filter

Unlike the white filter, the details of the sun captured by the Hα filter are very rich, even in places far away from sunspots. The most difficult place to tune is at the edge, because there are both details of the sun's surface and details that highlight the sun's surface (prominence). If the sun picture you take is close to saturation, that is, only a few pixels reach pure white, then it is possible to pull out the prominence hidden in the background by adjusting the color level.

There are usually two peaks in the color scale of this photo. The peak very close to black is the sky background, but it may also contain details of prominence. Another peak is the details of the sun.

Make a copy of the image and process it in two parts. For the first color scale, pull the middle slider to the left, which will highlight the outstanding details. The second one is used to adjust the surface of the sun and drag the middle slider to the middle of the main peak. The specific location depends on the effect, as long as the contrast of the edge is not adjusted too strongly, it will look natural.

Now, on this picture of the sun, choose a black background. Then enlarge the selection by a few pixels. Feather a few pixels to make the edge transition. Invert the selection, copy and paste it on the highlighted image, and adjust the position. In this way, you get a picture with both details of the sun and prominence.

Color processing-the biggest problem in planetary image processing

In all the steps of image processing, color processing is the most controversial. If you give 10 people the same images of red, green and blue planets and let them synthesize color images, you will definitely get 10 completely different results. This is because everyone has a different color of the planet. We always tend to adjust the color according to our subjective impression when dealing with it. It is easy for our brains to think that the color in this impression is right, even if it is far away from it.

The only way is to set a series of reference objects, process them according to the reference objects, and discard the colors in the brain. Easier said than done, even experienced people will inevitably make mistakes.

The colors of the sun and the moon are easy to handle, because the color of the moon is single, and everyone's subjective impression of the color of the sun is similar. This refers to images taken with white light and special filters, as well as color photos synthesized by traditional methods, rather than the real color of the sun.

However, on the planet, the situation is different, and it must be carefully adjusted to get the correct color. If possible, you should use a color camera and use automatic white balance, and the shooting result is correct. If the planet is too dark, increase the gain of the control software.

Color (red, green and blue) filters will be used for finer images. Pay attention to the same saturation when shooting, otherwise the result will be colored. See if there is this function in your control software.

Let's take it step by step:

Step 1: calibrate the monitor.

Before adjusting the color of the planet, you should check whether your monitor shows the color correctly. Many websites can calibrate the monitor online or use the software that comes with the monitor.

Step 2: Image Creation

Combine RGB color images with red, green and blue images. If the graininess of the result is too obvious, use "speckle removal" or slight Gaussian blur to remove noise. Don't be too vague, or the details will be lost.

Step 3: Curve

Copy the RGB layer and use the original RGB result as a backup. Open the curve tool and add adjustment points at the quarter. Adjust the overall brightness with the middle adjustment point, and then fine-tune the adjustment points on both sides according to your own preferences.

Step 4: Sharpen

Sharpening using Qualcomm filtering technique (discussed below). Qualcomm filter radius, slowly pull from minimum to maximum, and stop when the preview effect is the best. If the color fades, increase the color saturation.

Optional color

Flatten the layer and copy a new one. For Jupiter and Mars, use the optional color tools mentioned below to check. Select yellow and fine-tune the black slider to highlight hidden details.

Step 6: Compare

See if you have any color cast. The polar cap of Mars should be grayish white, the stripes of Jupiter are grayish orange, and the equatorial belt is whiter. Saturn's rings are gray and the equatorial region is dark straw yellow.

Color adjustment tool:

Color balance

Color balance is used to adjust the highlights, shadows and midtones of an image. When there is color deviation, such as too yellow or too red, adjusting color balance can eliminate color deviation.

Tone, saturation and brightness

The hue, saturation, and brightness level tools can adjust an image to a specific color. For different color ranges when hue is adjusted, the purity of color is adjusted by saturation, and the lightness of lightness level color (please refer to my translation of "6-color synthetic true color-multicolor synthetic technology of astronomical photos" for specific definitions and examples).

Optional color (image/adjustment/optional color)

It changes the amount of color in the basic color range. Originally used to calibrate the color of printed images, it is also useful to extract the details of planetary images. Choose red or yellow and adjust the black slider, which is often used when dealing with Mars and Jupiter.

Replace Color (Image/Adjust/Replace Color)

It can click on a specific color to adjust, and then adjust the hue, saturation and brightness of the color. This method is suitable for reducing the influence of unwanted colors.

Enhance the details of the planet —— Six ways to get perfect photos of the planet

high-pass filter

Qualcomm filtering can adjust the clarity of the image. This word is translated through a high-pass filter. I'm sure anyone with an engineering background will translate it like this, but it took me a long time to find a translation of "Gao Fancha Reservation" in my Photoshop. Please pay attention when looking in the menu filter/other. First, copy the layer where the image is located, and set the blending mode of the above layer to "Overlay". Select Menu Filter/Other/Gao Fancha Reserve, the size of the filter is defined by the radius, and the result of real-time adjustment can be previewed in PS.

The degree of sharpening can be well controlled by changing the value of radius. Usually start from a small value and increase until you are satisfied. Because the sharpening layer is on it, changing the opacity of the layer can also control the sharpening effect.

Superimposed inversion and blurring

The overlay blending mode combined with inversion and blurring operation can enhance the dynamic range of the image. This method is most effective for extracting the fuzzy and dark details of the moon and the gray shadow of the moon sea.

First, copy the layer. Select Menu Image/Adjust/Invert in the upper layer, set the opacity to 50%, and do Gaussian blur again. Adjust the degree of blur according to the effect. When finished, the image may look a little flat. Fine-tuning the contrast and reducing the brightness can restore the original overall visual effect of the image, but the details come out.

Optional color

Optional color (image/adjustment/optional color) tools originated in the field of printing. In the field of astrophotography, it can be used to deal with a specific color without changing the balance of other colors.

For example, the two belts near Jupiter's equator may contain red and yellow details, which are not easy to detect. Select yellow in the optional color tool, and then adjust the color slider below to enhance the color and show hidden details. Be careful not to adjust too much.

USM sharpening

USM Sharpening (Filter/Sharpening /USM Sharpening) is a common sharpening tool. It has three adjustment variables, "quantity" adjusts the intensity of contrast and "radius" adjusts the size of sharpening. These two parameters cooperate with each other, one is reduced and the other is increased, and the effect is similar.

The third parameter, called "threshold", defines the minimum difference that adjacent pixels must achieve. Usually 0, which is useful when there is a lot of subtle noise in the image. By setting the threshold slightly larger, USM sharpening can be avoided in these areas.

USM sharpening is addictive because the image enhancement is too obvious, but enough is enough. Give a good combination, and the three parameter combinations are: 50, 0.5, 0.

Shadow-highlight

Image/adjustment/shadow-highlight, which is best at dealing with dark parts and highlights without affecting the overall effect of the image. Its adjustment mechanism is similar to the above overlay anti-blurring technology, and other hidden gray details can be extracted.

The shadow-highlight tool is very useful for dealing with the hidden details in the moon photos and the details of the sun's edge hidden in the black background. It can also make up for the unclear edge of the planet that is not handled well.

There are two modes in Photoshop, Basic and Advanced (by checking Show More Options). The basic mode only provides two adjustment variables: shadow and highlight. Advanced mode has more options: quantity, hue width and radius.

Brightness (or brightness, luminosity, original luminosity)

The brightness of the image defines the spatial details of the image. The combination of lightness and color produces an image with both details and colors. Paste the sharpened image into the layer above the original RGB image, sharpen this layer as much as possible, and set the layer blending mode to "lightness" to see if the effect is very different. R-band images are usually used as luminance images to synthesize so-called R-RGB images. R- indicates that red is used as the brightness enhancement layer. Longer wavelength bands and infrared images usually look clearer and have Gao Fancha, so they are used as brightness images.

Brightness layer can change the color balance, so the composite image must be explained when it is used for scientific purposes. G-band brightness images can also be used under very ideal brightness conditions.