Page 2 for me, I've set my forum display to put 20 posts to the page.

@Pseudoephedrine:

Recognize this image?
click me

Worked about 3 weeks on deleting the grid. There are still a few artifacts left and some of the lines merged with the grid. But that can be handled in a later step.

So far it works only on scanned graph paper. No hex, no photographs. Those will come later.

Cool.

Hello again.

It's been some time since my last post here. I've had a lot of work with the thesis and i wanted to show some of the early fruits of it.

Check out these images:

First the original from Pseudoephedrine (thanks again!):
orignal

As you have seen before: the image without the graph paper lines:
degridded

Analysed Colors, smoothed the objects a bit and colored the numbers, so the programm knows it's not part of the walls:
colors

Finnaly vectorized, with some error correction (notice that some lines are complete where there where openings) and alligned to the grid, that has been extracted in the first step
final

It's of course not perfect. Some doors are not detected because of the drawing not being exact enough or because they were to close to each other. Also the stairs are not detected and it would be very hard to do so because the spacing between the lines is too small.

Since the final map is vectorized (apart from the numbers), there are a lot of different possibilties to display the map. The image above is just an basic example.

Btw. @Pseudoephedrine: What is the meaning of the S-shaped lines at some Walls (e.g. to the upper right of the 9)?

I hope you all like the vectorized map, because I have to ask again for a favor: I really need more maps for testing, so I can figure out how to improve my algorithms and to work out constraints.
Also I need more maps, because it's quite important for the evalutation part of my thesis.

Maps should be very basic (Lines and text only, like the one above). Colors are not important. Graph paper is ok. But I have not implemented the grid removal for hexpaper, yet.

So if you have maps like that, I'd greatly appreciate if you scan them and send them to me (You can upload them to 666kb.com or similar sites and post the link here or send it to me via pm).

Karto

I will try to send you a couple, but note that there are a bunch of hand-drawn maps, reasonably scanned, already available at plagmada.org.

Quote from: stu2000;514675I will try to send you a couple, but note that there are a bunch of hand-drawn maps, reasonably scanned, already available at plagmada.org.

Wow, thanks a lot! This is exactly what i need. Didn't know there where archives like that. If you know any more sites like that, I'd appreciate links. The more testing material, the better.

The "S" through a wall or door means its a secret door.

Also, Pseudo dosn't post here anymore, FYI, be got ban-hammered.

Quote from: Karto;482117Now i need your opinion:
- How do you like the idea?
- Would you use a tool like this?
- What features does it need?
- What features would be nice to have?
- If needed: would you mind some (hopefully minor or non existent) restrictions on how you draw your maps?

This is some very interesting stuff.

(1) I would almost certainly use this tool if you got it hooked to an online gaming table. One of the biggest problems I have with online gaming tables is the amount of extra prep work it takes to digitally prep files. If it was as simple as "scan my map, tweak the output for 5 minutes, and then I can play" that would make a huge difference.

(2) I would say the biggest feature would be support for light sources. Both defining light sources the PCs are carrying and light sources in the dungeon. These would both need to be update-able during play.

Thanks, Werekoala.

Quote from: Justin Alexander;515056This is some very interesting stuff.

(1) I would almost certainly use this tool if you got it hooked to an online gaming table. One of the biggest problems I have with online gaming tables is the amount of extra prep work it takes to digitally prep files. If it was as simple as "scan my map, tweak the output for 5 minutes, and then I can play" that would make a huge difference.

(2) I would say the biggest feature would be support for light sources. Both defining light sources the PCs are carrying and light sources in the dungeon. These would both need to be update-able during play.

I'll keep that in mind. Both points shouldn't be too hard to implement. Just might take a while.

For now it's the analysis of the map that is important for my thesis. Today i plan to start actually writing stuff (instead of coding). Although I'm pretty lazy about it and would prefer to continue coding .

Once the thesis is over I'll start working on the final application, which in itself will take quite a lot of time. I'm even thinking of getting a few other people to help me code.