Project 5: Cover Photos
The purpose of this project is to make a program that processes a list defining a collage and builds the resulting image. The collage list information can specify if the program should apply an effect and/or an alpha blend to each image.
Here is the reference guide to the Zelle graphics package.
For this assignment you're going to create a few new functions. The getImageSize and buildCollage functions should go in your collage.py file and will process the collage list information and build the output image. The placeImageNoBkg function should go in your filter.py file and will be identical to placeImage, except that it will not transfer blue/green screen pixels into the destination image.
Given the collage list structure we developed, write a function that calculates
how big the program needs to make the background Image to hold the collage.
The strategy we outline below is to have a variable (rows) that keeps track of the biggest number we see in the collage list that represents the bottom row of an image in the collage. I.e. if we have 2 images and one of them has a y-offset of 100 and is 300 pixels tall, then that image tells us we need at least 400 rows of pixels in our collage. If the second image is at a y-offset of 0 and is 500 pixels tall, then that image tells us we need at least 500 rows of pixels in our collage. Since 500 is bigger than 400, rows gets to be 500. We loop through every image, calcuting this bottom-row number and update the rows variable is the new bottom-row number is bigger than the value rows currently has.
We use an analogous strategy for the columns.
def getImageSize( clist ): # assign to the variable rows, 0 # assign to the variable cols, 0 # for each item in clist # assign to x0 the x offset information in item # assign to y0 the y offset information in item # assign to src the Image reference (last element in item) # assign to dx the offset x0 plus the width of src # if dx is greater than cols # assign to cols, dx # assign to dy the offset y0 plus the height of src # if dy is greater than rows # assign to rows, dy # return a tuple with cols and rows in it
Test out your getImageSize function using the simple collage list above, after reading in the images using your readImages function. The answer should be 600 wide (cols) and 450 high (rows) if you use maine1.ppm and main2.ppm as test images.
Now that you have practice working with the collage information lists, you should be able to write a generic buildCollage program that creates a collage from the information in the list. And this is precisely what you will do in the project.
We will be adding a new item to the collage list to specify whether
the blue/green background should be removed when placing the image
into the collage. Therefore, instead of 6 items in each sub-list,
there will now be 7. You will need to add another IDX type constant to
your set of indexes.
You can use whatever ordering of information you like, but the provided test code assumes that the remove background information is at index 5 and that the Image information will be at index 6.
In your collage.py file, create a function buildCollage, as given below. Unless otherwise specified, each comment is a line of python code in the final function. The indentation of the comments should match the code.
def buildCollage( clist ): # assign to (cols, rows) the result of calling getImageSize with clist # assign to dst the result of calling graphics.Image with a Point of (0,0) and of size cols and rows # for each item in clist # assign to x0 the X-offset element of item (i.e. use the IDXXOffset index) # assign to y0 the Y-offset element of item # assign to operator the filter element of item # assign to alpha the alpha element of item # assign to noBkg the remove background element of item # assign to src the Image element in item # use a set of if/elif statements to apply the correct # operator to the image (this is several lines of code) # call placeImage to put src into dst at location x0, y0 with blend alpha # return dst
You can use the function testbuildcollage.py to test your buildCollage function. Read through the test function before you use it. The test function assumes that your build collage knows what to do with the operator strings 'rbswap' and 'original', as we went over in lab.
As always, remember to remove the comments we supply after you have written your code.
- In your filter.py file, create a function placeImageNoBkg that has the same arguments and functionality as placeImage, but does not copy the src pixel into the destination image if the src pixel is part of the background screen. Your function needs to work only for one of the background screens (blue or green). A good extension is to have your function handle both, possibly by adding a parameter.
- Edit your buildCollage function so that it uses the remove background information in the collage list, stored in the noBkg variable, to determine whether to call placeImage or placeImageNoBkg as the last step in the main for loop.
Make a new python file (e.g. mycollage.py) that contains a single main
function. Be sure to put a call to the main function behind a test on
if __name__ == "__main__": main()
The main function should create a collage information list, call readImages, call buildCollage, then write the image to a file. You can use the testbuildcollage.py file as an example. Your collage should use at least four different source images.
The collage should include at least five different sub-images. Some of these can be copies of one of the source images with different effects. The collage should use at least 3 effects, at least one alpha blend (alpha < 1), and at least one blue-screen or green-screen image where the background pixels are not copied into the collage. You can access your photos here
Required image 1 is your first collage.
Make a new python file, coverphoto.py, that takes in at least one
image filename from the command line. It should build a collage
appropriate for a FB cover photo and write it out to a file.
The collage should be about three times as wide as it is tall and have at least three sub-images. You can control the height and width of the collage by appropriate selection and placement of your images, or you can fix the relative height and width and change placeImage to safely handle images that go outside the image boundaries. You can choose random effects for the three images or pick specific effects. The collage should use at least two effects.
Required image 2 is your cover photo collage.
These are just examples, not necessarily recommendations. Please feel welcome and encouraged to design your own. A great extension genuinely interests you, pushes your understanding of CS, and inspires you to learn something new.
- Create some new image effects you didn't implement in the last lab.
- Enable the placeImageNoBkg function to handle both the blue and green screens. You can try to do this automatically (e.g. by checking some part of the image) or by adding a parameter that indicates which kind of background screen is being used.
- Make some additional programs that generate collages with specific geometric arrangements and that take their list of image files from the command line.
- Add additional options to the collage information list, like whether to mirror the image or not, in addition to the other options.
- Create a collage dynamically based on how many images are given on the command line.
- Figure out how to call the proper effect function on each image without using an if/elif control structure. (Ok, this is kind of esoteric, but it's a cool feature of python.)
- Try scaling an image when you put it into the collage. If you want, stick to powers of two.
- Uber-extension: put the collage information in a text file. Have your program read the text file (given as a command-line argument), parse the information, and build the collage. For this to be an uber-extension, you can't write the text file as python code.
Hand-in and Report
- Put the python files you wrote on the Courses server in your private directory in a folder named Project5.
- Make a new wiki page for your assignment. Put the label cs151f18project5 on the page.
In general, your report should follow the outline below.
- Title includes your name and clearly describes the project.
- Section headings are used to delineate distinct sections of the report.
- Abstract identifies key lecture concepts (e.g. code structres, data types, and libraries) relevant to the project.
- Abstract explains why key lecture concepts are important to achieving project goals.
- Abstract identifies program output(s), giving context to the project tasks.
- Solutions to tasks are described, focusing on how you used key lecture concepts to solve each task.
- Required images/outputs are present and clearly labeled.
- Reflection at the end of the report addresses how the lecture concepts mentioned in the abstract made this project possible. If you can think of a more elegant way to achieve the same results, please share!
- Sources, imported libraries, and collaborators are cited, or a note is included indicating that none were referenced.
- Don't forget to label your writeup so that it is easy for others to find. For this project, use cs151f18project5