Talk:Bitmap/Flood fill: Difference between revisions

From Rosetta Code
Content added Content deleted
(Prereq)
(proposal for 3d format)
Line 5: Line 5:


How about a 3d flood fill?<br>
How about a 3d flood fill?<br>
the scipy subpackage ndimage has a multidimensional implementation for reference --[[User:tinku99|tinku99]][[User:Tinku99|Tinku99]] 01:26, 26 November 2009 (UTC)
the scipy subpackage ndimage has a multidimensional implementation for reference --[[User:Tinku99|Tinku99]] 01:26, 26 November 2009 (UTC)
: Interesting, but I think we'd need a common volumetric target format, similar to [[Basic bitmap storage]], but in three (or, more interestingly, N.) dimensions. I'm not saying that's hard, but it's something of a prerequisite. ( a subsequent derivation of the image format to support M channels per pixel would be even more interesting.) --[[User:Short Circuit|Michael Mol]] 08:38, 26 November 2009 (UTC)
: Interesting, but I think we'd need a common volumetric target format, similar to [[Basic bitmap storage]], but in three (or, more interestingly, N.) dimensions. I'm not saying that's hard, but it's something of a prerequisite. ( a subsequent derivation of the image format to support M channels per pixel would be even more interesting.) --[[User:Short Circuit|Michael Mol]] 08:38, 26 November 2009 (UTC)

IMO the easiest thing would be to use netpbm pgm and ppm formats and simply allow more than 2 dimensions on the line where we normally puts rows and columns. alternatively we could have a simple binary format where the first line would be the equivalent declaration in c followed by a newline character and then the binary data of the image / matrix [[User:Tinku99|Tinku99]] 07:21, 29 November 2009 (UTC)

Revision as of 07:21, 29 November 2009

What about ppm image? Since we already have read ppm file task. --Dmitry-kazakov 15:13, 18 February 2009 (UTC)

Ops, yes, I've used Read image file through a pipe, which has just 1 C implementation (mine:D). Now I am leaving, but tonight I will upload the ppm converted image, if no one else already did that. --ShinTakezou 17:51, 18 February 2009 (UTC)
Back to my steps: I can't upload PPM files; tried cheating with a different ext, but it says "corrupted"... so... it is up to you to convert the jpg into PPM if you need it... ;) --ShinTakezou 22:27, 18 February 2009 (UTC)
In case someone can't convert it, you can download a 50% smaller PPM version here (until the site will be up :D)--ShinTakezou 00:59, 19 February 2009 (UTC)

How about a 3d flood fill?
the scipy subpackage ndimage has a multidimensional implementation for reference --Tinku99 01:26, 26 November 2009 (UTC)

Interesting, but I think we'd need a common volumetric target format, similar to Basic bitmap storage, but in three (or, more interestingly, N.) dimensions. I'm not saying that's hard, but it's something of a prerequisite. ( a subsequent derivation of the image format to support M channels per pixel would be even more interesting.) --Michael Mol 08:38, 26 November 2009 (UTC)

IMO the easiest thing would be to use netpbm pgm and ppm formats and simply allow more than 2 dimensions on the line where we normally puts rows and columns. alternatively we could have a simple binary format where the first line would be the equivalent declaration in c followed by a newline character and then the binary data of the image / matrix Tinku99 07:21, 29 November 2009 (UTC)