Talk:Zhang-Suen thinning algorithm
Inverted axis?
Could someone help explain why, in the example image the following transformation occurs (in the periods separating the R and C, and after the C):
..... ..... .###. -> ..... .#?#. ..#.. ..... .....
Surely the the cell labelled '?' will be culled at step 1:
- It is black with 8 neighbours - B = 5 (2 <= 5 <= 6) - A = 1 - At least one of P2 P4 P6 is white (P6 is white) - At least one of P4 P6 P8 is white (P6 is white)
Why isn't it whitened at step 1?
--Tim-brown (talk) 17:33, 15 October 2013 (UTC)
- I expect that there are removals of some of those surrounding cells before it gets to your '?' cell which affects the final outcome. --Paddy3118 (talk) 18:54, 15 October 2013 (UTC)
- I'm considering what happens to this individual cell (not necessarily its neighbours). As far as I am concerned, the situation I describe above is the calculation for Step-1 of the first iteration. Nothing has changed (i.e. been removed) before this: all changes are stored, and applied `after` the analysis. So the step-1 rule should apply to cell '?'. And it should be blank (by my interpretation of the rules). But it ain't.
--Tim-brown (talk) 19:27, 15 October 2013 (UTC)
You are right the table given of P1 to P9 in the task page is actually transformed but ins such a way that the output is thinned in the same way (but with an offset possibly). I am actually calculating with:
| P7 | P6 | P5 |
| P8 | P1 | P4 |
| P9 | P2 | P3 |
I've got an inverted vertical axis it seems. --Paddy3118 (talk) 22:32, 15 October 2013 (UTC)
- Now I'm a bit more confident in posting my Racket solution! ---Tim-brown (talk) 09:50, 16 October 2013 (UTC)
Is the thinned version of the big "R.C." in the task description correct? I have it as:
........................................................... ........................................................... ....#.##########.......................#######............. .....##........#...................####.......#............ .....#..........#.................##....................... .....#..........#................#......................... .....#..........#................#......................... .....#..........#................#......................... .....############...............#.......................... .....#..........#...............#.......................... .....#..........#................#......................... .....#..........#................#......................... .....#..........#................#......................... .....#............................##....................... .....#.............................############............ .......................###..........................###.... ........................................................... ...........................................................
---Tim-brown (talk) 10:05, 16 October 2013 (UTC)
Three Data Files
To avoid every entry to needlessly contain the same large tables, I suggest to add to the task description three links to the tree files of the test cases. -bearophile (talk)
- I think if the outputs are typically in ./# form, the inputs probably should be too, rather than csv. --TimToady (talk) 15:53, 21 October 2013 (UTC)
Mathematica's in-built algorithm
I converted Mathematica's ones and zeros to something I could see and it seems that their in-built thinning algorithm does not give the same result as that of Zhang Suen.
I think that it is right and proper to show the output of the in-built function as I would expect that to be used, but I also asked for an implementation of the ZS algorithm as I maked it incomplete.
If you think that is wrong then just give your argument please.
The following is just a code dump of how I converted to ascii art for comparison using Python:
<lang python>>>> print( { {0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0}, {0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0}, {0, 1, 1, 1, 0, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0, 0, 1, 1, 1, 1, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0}, {0, 1, 1, 1, 0, 0, 0, 0, 1, 1, 1, 0, 0, 0, 0, 0, 1, 1, 1, 0, 0, 0, 0, 1, 1, 1, 0, 0, 0, 0, 0, 0}, {0, 1, 1, 1, 0, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0, 0, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0}, {0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0}, {0, 1, 1, 1, 0, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 0, 0, 0, 0, 1, 1, 1, 0, 0, 0, 0, 0, 0}, {0, 1, 1, 1, 0, 0, 1, 1, 1, 1, 0, 0, 1, 1, 1, 0, 1, 1, 1, 1, 0, 0, 1, 1, 1, 1, 0, 1, 1, 1, 0, 0}, {0, 1, 1, 1, 0, 0, 0, 1, 1, 1, 1, 0, 1, 1, 1, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 1, 1, 1, 0, 0}, {0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0}}.replace('{', '(').replace('}', ')'))
( (0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0), (0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0), (0, 1, 1, 1, 0, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0, 0, 1, 1, 1, 1, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0), (0, 1, 1, 1, 0, 0, 0, 0, 1, 1, 1, 0, 0, 0, 0, 0, 1, 1, 1, 0, 0, 0, 0, 1, 1, 1, 0, 0, 0, 0, 0, 0), (0, 1, 1, 1, 0, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0, 0, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0), (0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0), (0, 1, 1, 1, 0, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 0, 0, 0, 0, 1, 1, 1, 0, 0, 0, 0, 0, 0), (0, 1, 1, 1, 0, 0, 1, 1, 1, 1, 0, 0, 1, 1, 1, 0, 1, 1, 1, 1, 0, 0, 1, 1, 1, 1, 0, 1, 1, 1, 0, 0), (0, 1, 1, 1, 0, 0, 0, 1, 1, 1, 1, 0, 1, 1, 1, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 1, 1, 1, 0, 0), (0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0)) >>> inp = ( (0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0), (0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0), (0, 1, 1, 1, 0, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0, 0, 1, 1, 1, 1, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0), (0, 1, 1, 1, 0, 0, 0, 0, 1, 1, 1, 0, 0, 0, 0, 0, 1, 1, 1, 0, 0, 0, 0, 1, 1, 1, 0, 0, 0, 0, 0, 0), (0, 1, 1, 1, 0, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0, 0, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0), (0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0), (0, 1, 1, 1, 0, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 0, 0, 0, 0, 1, 1, 1, 0, 0, 0, 0, 0, 0), (0, 1, 1, 1, 0, 0, 1, 1, 1, 1, 0, 0, 1, 1, 1, 0, 1, 1, 1, 1, 0, 0, 1, 1, 1, 1, 0, 1, 1, 1, 0, 0), (0, 1, 1, 1, 0, 0, 0, 1, 1, 1, 1, 0, 1, 1, 1, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 1, 1, 1, 0, 0), (0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0)) >>> print('\n'.join(.join(('#' if x else ' ') for x in line) for line in inp))
######### ########
### #### #### ####
### ### ### ###
### #### ###
######### ###
### #### ### ###
### #### ### #### #### ###
### #### ### ######## ###
>>> print('\n'.join(.join(('#' if x else ' ') for x in line) for line in out))
#### ###
# # # ###
# # #
# # #
## # #
# ### #
# # # ##
## ## #### ##