Introduction to Image Processing

This project introduces the students to the art of image processing. They will learn how to adjust astronomical images to bring out interesting features. Students will also learn how to make color images like the ones displayed on SkyServer. Students will experiment with creating color images using different filters to see what other features show up. They will also learn how to "blink" images, a technique that can be used to find asteroids, supernovae, or other changes in images.

References

Iris Software Homepage

Introduction to Astronomical Image Processing, by Richard Berry

Project Goals

By the end of the project, students should be able to:

• Retrieve raw images from the SDSS database
• Display images using IRIS, a freeware image processing program
• Adjust maximum and minimum brightness to bring out desired features in an image
• Create tricolor images
• Describe the differences between an image created with various filter combinations, such as ugr and gri
• Align two separate images by shifting them until corresponding stars are aligned
• Find changes in images by using the "blink" command
• Apply different color scales
• Understand log scaling and use it when appropriate
• Make a 3D view of an image

Background Knowledge

Students should know the uses of color filters in astronomy. The Colors unit provides a good introduction to filters. Students should also be familiar with color mixing in light, and with how three grayscale images can be combined to produce a color image.

Computer skills are a major component of this project. Students should know how to download files from the Internet and unzip them using a program such as Winzip. They will need to know how to create directories in Windows and how to move their files into appropriate directories. Students also will need to know how to rename files (since the default file names that SDSS uses are long and cumbersome). Computer skills will probably be a major stumbling block to some students; you may wish to offer an optional review session to teach these skills before starting the project.

Project Structure

The first page gives background information on how a CCD image is displayed using image processing software. This process is used by all imaging software and is not specific to Iris. You might wish to print out the table of numbers and have your students fill it in with their own color palettes. Tell them to use fewer colors, say four. They would then have to make choices (for instance, one color could represent 0-2, another color could represent 3-5, etc.) You could use this exercise to illustrate how different image representations of the same data can convey different information. This extra exercise would add about 10-15 minutes for the students, plus another 10-15 minutes of discussion. You might have students post their diagrams so they can compare with other groups. Emphasize that the SDSS produces five of the images for each part of the sky, one for each of our filters.

The next section introduces students to Iris, the image processing software they will use in this project. If you have time, you might download and install the software for the students in advance. Iris is freeware. Use of the software is free, but you may not decompile or modify the source code.

The most important task for students in this section is to set up a directory for their images and tell Iris what directory they are using. If the students use a different computer the next day, they may need to set the directory again. This section should not take more than 5-10 minutes.

Next, students will download their first images. The first set of images are atlas images, small cutouts of a specific object. The images are fairly small files and download fairly quickly. This section should not take more than a couple of minutes.

Next, the students will look at the max and min features of the software. These features allow the students to focus on bright areas of the image or dark areas of the image. This exploration will probably take 10 - 15 minutes.

Next, the students will learn to make a tricolor image by combining combine the g, r, and i images. The atlas images are already aligned, so the students will not have to worry about aligning them for this section. They will attempt to reproduce an image in our Famous Places. Allow about 15 minutes for this section.

The students will then make tricolor images using other filters to see what other features stand out. Allow about 20 minutes for them to create and explore these images.

Next, they will try to find asteroids by blinking two images. Students will download corrected frames for this section. These files are large and take longer to download (you may retrieve them in advance for your students if you have a slow internet connection). The students will have to align the frames before blinking them. They will then try to find asteroids in a randomly selected frame. Allow about 45 minutes for finding asteroids.

Students will then explore how to apply log scaling and color palettes to their images. Allow about 20 minutes for this section.

Next, they will learn alternate ways of viewing their images, such as isophotes (lines of constant brightness) and 3D views. This section should take 20-25 minutes.

This exercise has two choices for a final challenge. The final challenge is designed to let students conduct independent self-directed research. It should not be done in the classroom, but you may wish to offer extra credit for completing it. Students can make tricolor images using 2MASS data, or they can look for "i dropout candidates" in SDSS data.

Questions and Exercises

Questions are designed to get students thinking about the way scientists work. Exercises are designed to get students to explore SkyServer data. For answers to all questions and exercises, email us at helpdesk@sdss.org.

Students should be evaluated based on their written answers to the questions and exercises. You may use our sample scoring rubric or develop your own. If you use our scoring rubric, print out a copy for each student and attach it when you return his or her work.

For specific information on any part of the project, click Next.