Grading PIL Projects
As mentioned before, there are two main use cases for PIL projects - these are projects that have a lot of images, and projects that do heavy image manipulation. When PIL is used to simply display images or apply basic image manipulation, such as applying contrast to an image, this isn't really considered to be added complexity, as its just making basic function calls and instantiating Image objects. In fact, this type of image manipulation can even be considered to be less complex to do in PIL than with just Tkinter, since images in Tkinter are just very hard to use.
With more complex sounding techniques, such as edge detection etc., it all depends on whether or not you implement it on your own. Most things can be implemented with just a couple lines of PIL code - in this case, it would not be as complex, since these algorithms are readily available online. Implementing it by yourself though requires a solid understanding of how these algorithms work on the backend, so this should be considered as sufficient complexity.