JoeJ said:

… almost demystified :D

Two questions remain:

You mentioned a training process in the paper, which misses from your given list of processing. I was assuming there is some NN stuff going on, so that's maybe another source of confusion.

It's not really a NN. It's more like an SVM or KNN instead, but you are right that it's all about classification. As for the word ‘training’, we train in a bunch of batches and then finally use the algorithm to classify a test point.

But i still wonder how you get them and they look interesting. Now i guess it could work like this:
For a given location, sum up the distances to all points, also track the closest point.
Search for spots where the distance to the closest point is 10%, 20%, 30%… from the sum/average of all distances, and draw them so we get a loop around the closest point, which is not just a perfect circle.

Regarding figures 9 - 14, what happens is that I train using 50 batches. Each batch produces an image, which I use to create a line segment / triangle set. These line segments are what are drawn in transparent grey.

After all 50 batches are computed, the 50 images are averaged into one image. This average image is what's used to create the pure black ‘average’ line segments.

I do not really intend to help out on that. Knowing nothing about statistics i miss any background knowledge, and i'm not the kind of person which would read the paper.
So i can not really seriously review the paper. Just discussing out of personal curiosity.
But you know that ; )

Thanks anyway though!