haskell-photoshop

Names: Jason Yan and Justin Zhang Pennkeys: jasyan and judtin

We created Haskell Photoshop, which processes images and performs various manipulations on them. A key component of our photoshop program is our extensible design that allows multiple implementations of the same image processing functions using different image representations. Since they all implement the same functions, a user can pick and choose an implementation that is best suited for their task. Becuase of our multiple implementations, we put an emphasis on rigorous testing and profiling of them.

Module organization

I would reccomend start reading from the (short) app/Main.hs or app/Benchmarking.hs in order to get a sense of how the image processing functions are used. Then, move onto src/PPM.hs, our simpler implementation, before reading src/QuadTree.hs.

app/Main.hs: Home of our main function, uses supplied by a image representation file in order to manipulate images at a high level.

app/Benchmarking.hs: Runs a benchmark suite on the image processing functions available, timing their runtimes. Run suite with stack bench.

src/PPM.hs: An implementation of our image processing functions using the PPM (Portable Pixel Map) format, which is just a 2D array of pixels.

src/QuadTree.hs: An implementation of our image processing functions using the QuadTree format. The QuadTree fully supports rectangular images. In addition to general functions listed below, supports an extra lossyCompression function.

Image Processing Functions

Here are the functions supported by all image representations

readInput toJpg toPng

blur changeColor crop grayscale reflectHorizontal reflectVertical rotateLeft rotateRight saturate