Studying Fuseki/go
Posted on 2003-12-29 by ivo :: /go :: link
Now that I'm a little further in mastering the game, I can see
where my weaknesses lie and what I should do or which books I
should read to get more skilled. Right now I think that my
greatest weakness lies in the opening, the 布石. So I went to the
store and bought a few books about it. Reading these books makes
me aware that yes – I needed it.
There are lots of areas that I need to master, but the fuseki was
the most urgent. I think the next field of interest should be the
endgame, but I'm not sure yet. Reading these books on the fuseki
pushed me up quite a bit, and I haven't even finished
yet…
But even when studying these exotic games the books give me, I
still feel that I'm picking up on the principles that lie behind
them, even though I'm not always sure about what it is.
It is the experience of suddenly seeing yourself play a new type
of move, because you saw the principle behind it explained
somewhere, that makes you feel good. It doesn't even matter if
you lost or won the game as long as you can see you've learned
something.
Kifu recognition (part 2)/go
Posted on 2003-12-08 by ivo :: /go :: link
Well, the first steps have been made, I can now find lines in an
image. Below you will find an example of what sort of result this
yields. Unfortunately this is only the easiest part of step 1,
recognizing the grid on the goban… :(

Here are some of the
intermediate steps that were necessary to get this image. I put
another example here.
Kifu recognition/go
Posted on 2003-12-06 by ivo :: /go :: link
For a while now I've been thinking about writing a program that
can take an image of a go board and translates that to a kifu in SGF. I know this has been
done
before, but as far as I could see, there isn't really a
readily-available program to do this, let alone any free
software.
So I'm accepting the challenge. Here's a simple list of a few of
the things this program has to be able to recognize, roughly in
increasing difficulty:
- The grid on the board;
- The stones on the board and their exact positions on the
grid;
- The color of the stones;
- The size of the board (19×19, 13×13, 9×9);
- Any markings on stones;
- Number/letter markings on stones;
- Number/letter markings anywhere on the board;
- The position in a photograph of an actual game, where stones
may not be placed exactly on the points in the grid.
Starting with generated images and scans of images in books, I
will be working my way down this list. I'm now trying to get a
simple program that does nothing more than recognizing lines on
the board, and define where the grid must be.