Hi Dr. Breuel

please correct me if I am wrong, but to be able to use RegionExtractor
first I have to set it to one of the following

regions:setPageColumns(seg)
regions:setPageParagraphs(seg)
regions:setPageLines(seg)

so if the segmentation function work correctly I should be able to
identify the (columns,paragraphs or lines) on the image.
I have uploaded an image "1CP.png", it is segmented by 1CP, however it
looks like the algorithm consider  the line as first priority rather
than columns, I am wondering is there something to do with the
implementation of the 1CP algorithm?

Also since the version 0.1.1 are there any update/fine tune on the
function of "make_SegmentPageByRAST()", "make_SegmentPageBy1CP()" and
"make_DeskewPageByRAST()"?
because I am still using version 0.1.1 and might consider to install
version 0.3 if those function has been changed, as sometime I find it
took long time to deskew a slightly large image, and sometime I will
get segmentation fault (suspect of running out of memory).

Regards,
Leo





On 10月21日, 上午7時22分, "Thomas Breuel" <[EMAIL PROTECTED]> wrote:
> Just count the number of distinct colors in whatever way you like: put the
> color values into a hash table, then count the keys, etc.
>
> A simple way is to call "renumber_labels" in iulib to renumber and count
> them for you; it returns the number of components.  RegionExtractor will
> also count them for you and let you iterate through them.
>
> Tom
>
> On Mon, Oct 20, 2008 at 12:35, Leo <[EMAIL PROTECTED]> wrote:
>
> > Hi All,
>
> > Just wondering is it possible to tell how many component has been
> > segmented after I segment an image use either RAST or 1CP
> > segmentation.
>
> > After I apply the segmentation if I apply simple_recolor()  to the
> > intarray it will apply different color to the segmented components on
> > the image, therefore I was wondering how can we tell how many
> > component has been segmented after the segmentation?
>
> > Regards,
> > Leo
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups 
"ocropus" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at 
http://groups.google.com/group/ocropus?hl=en
-~----------~----~----~----~------~----~------~--~---

Reply via email to