Held: Monday, 22 September 2014

Back to Outline 13 - Testing Your Procedures, Revisited. On to Outline 15 - Transforming Images.

Summary

Today we explore common ways in which to transform colors, pixels, and images.

Related Pages

• Reading: Transforming RGB Colors
• Lab: Transforming RGB Colors
• EBoard (Source) (HTML)

Overview

• Review: Color basics.
• Integer-encoded RGB colors.
• Computing new colors from old.
• An example.

Administrivia

• New lab partners!
• Quiz 3 returned.
• I think all of you saw my email about office hours. I'll be in as much as I can today to answer questions on the exam or to talk about difficulties (or both).
• "When do we have the final?"
  • From the home page: The final examination for this course is optional. It can be used as a makeup for one examination. I will count the final examination only if it is higher than your lowest examination, and it will replace the grade for that examination. Our final examination is scheduled for 9:00 a.m. on Friday, Dec. 19. However, since my other course has a final at 9:00 a.m. on Tuesday, December 16, I can also offer the final then. (If neither time works for you, let's talk. I'll find another time.)

Upcoming Work

• Lab writeup: Exercises 5e and 7b: http://bit.ly/151-2014F-w14
• Exam 1 due Tuesday night.
  • Bring printed copies on Wednesday.
• Reading for Tuesday:
  • Transforming Images

Warnings on Today's Lab

• In some instances, it may be difficult for some folks eyes eyes to distinguish certain colors. You may have to look at actual pixel values to see the difference.
• Sometimes Gimp doesn't like to show changes to an image. We haven't figured a way around that, other than clicking on the image (not just the window).

Extra Credit Opportunities

Academic

• Convocation Wednesday: Tara Zahra on Emigration from Eastern Europe
• CS Extras Thursday at 4:30: Mathematical Image Synthesis Toolkit
• CS Table Friday at noon: ???
• Campus Town Hall September 30

Peer Support

• Noteworthy, Sat., Sept. 27, 3:30 p.m. Herrick
• Science poster session, Sat., Sept. 27, morningish, science elbow
• Football Games (Saturday, Sept. 27, at 1pm)
• Men's Tennis (???)
• Women's Tennis (???)
• Noteworthy, Sat., Sept 27, 3:30 p.m. Herrick
• Anna Christie, Oct. 9-12 (SB plays Marthy)
• Evan's radio show 5pm Friday nights on KDIC
• Donna's radio show Sunday midnight on KDIC

Miscellaneous

• Participate in #GrinWell.
• Friends of Drake Library needs help setting up for their annual booksale on Thursday, October 2 (conveniently, the same night as Grinnell High School's homecoming parade). You can help a good cause and probably get a few free books, too.

Review: RGB Colors

• A standard way to represent transmissive colors.
• Think of it as a combination of three colored lights of varying intensities
  • Red
  • Green
  • Blue
• Why these colors? It just works that way.
• How do we measure intensities?
  
  • Some people like a 0-1 scale
  • Computers like a 0-255 scale (powers of two rock!)
• Important procedures:
  • (irgb-new red green blue) - create a new color
  • (image-set-pixel! image col row color) - set the color of a pixel
  • (image-get-pixel image col row) - get the color of a pixela
  • (irgb-red color) - get the red component
  • (irgb-green color) - get the green component
  • (irgb-blue color) - get the blue component
  • (color->rgb-list color) - get a list of the three components
  • (color->irgb color) - convert some representations to the RGB representation.

Transforming Colors

• Today we are exploring ways to transform colors.
• Why?
  • Deepens our understanding of colors.
  • We can apply a color transformation to a pixel in the image.
  • We can apply a color transformation to each pixel in an image. (More on that tomorrow.)

An Example

• MediaScript comes with a variety of built-in color transformations.
• However, we can also define our own (yay!).
• The normal framework for a color transformation is

(define color-transform
  (lambda (color)
    (irgb-new ___ ; computation of red component
              ___ ; computation of green component
              ___ ; computation of blue component
              )))

• Let's consider a simple one: We'll keep only the blue component of an image.
• How would we write this?

(define only-blue
  (lambda (color)
    (irgb-new ___
              ___
              ___)))

• Some tests
• While we'll spend today using this idea with pixels, for this example, we'll explore the effect on a sample image.