Code as a Liberal Art, Spring 2025

Unit 1 Project

Command line image generator

Assigned: Thursday, February 6

Due for in-class presentation, Thursday, February 27 (with 2-4 people presenting works-in-progress earlier on Tuesday, February 20) [NOTE NEW DATES]

Final draft of code: due in your Google Drive Folder on Friday, February 28, 8pm. (If you wish to submit the optional written assignment outline also due today and need to adjust one of the deadlines, please email me.)

Create a command line image generator that takes at least 4 command line arguments and uses those to modify and combine images into a digital collage.

These arguments should include at least two nouns, each with a corresponding modifier. Like this:

$ python3 generate.py <MODIFIER 1> <NOUN 1> <MODIFIER 2> <NOUN 2>

For example:

$ python3 generate.py colorful trees dark mountains

Your program should accept at least three possible values for the NOUN argument, each of which should each correspond to a folder of images from which your program randomly selects one or more images.

Your program should accept at least three possible values for the MODIFIER argument, each of which should correspond to a predefined filtering process that you apply to the corresponding specified image.

When run, your program will select an image (or images) for each NOUN, apply to the filter corresponding to the specificed MODIFIER, and then combine the filtered images together in some way.

If you receive any command line arguments that your program doesn't recognize, offer a helpful message and exit gracefully (i.e. by calling exit()). For example, "I don't understand CLOUDS"

Your user (or you when you are demonstrating your work) should be able to combine nouns and modifiers in arbitrary ways.

Process & tips

Start by making a new folder for this project. Within that, make one folder of data (images) for each noun that you plan to accept.

You can think about your modifiers in purely formal terms or something intended to invoke some meaning. For example, you might combine the images as a comic strip, a grid (uniform or for a page layout), torn posters on an urban wall, translucent layers, a digital glitch, a Dadaist collage, or a painting in the style of Cubism, Pointilism, Impressionism, or something else. Or you might play with figure/ground relationships to try to extract the figure from several images and collage them onto one background.

Think about all the image manipulation techniques we've talked about in class, and see the Python Image Library documentation for additional commands that you can use that we haven't spoken about yet. Think about how you might de-saturate an image to make it black and white (set the second component of pixels to 0 in HSV mode, or use convert() to change the image to black and white and save it), resize the images, copy/paste in fragments, make some or all of the image partly or fully transparent, or overlay some glitchy graphics.

For your nouns, you may think of concrete things like people, buildings, or birds, or even more abstract concepts or emotions like love, fear, or the color blue. I like to use Flickr and specify some search parameters:

---