This topic and its replies were posted before the current version of LEADTOOLS was released and may no longer be applicable.
#1
Posted
:
Wednesday, February 1, 2006 3:25:03 AM(UTC)
Groups: Registered
Posts: 17
Hi,
I'm using LT V14 Medical Version through Delphi 7 and I want to get all the light objects.
I use filters for pre-treatments and then, I binarize the image and I
have my light objects but I don't know how to have the informations of
each of them: areas, positions...
Does someone have a clue for me?
Thanks
Julien
#2
Posted
:
Thursday, February 2, 2006 1:42:23 PM(UTC)
Groups: Manager, Tech Support
Posts: 367
Was thanked: 1 time(s) in 1 post(s)
Julien,
If I understand correctly, you have multiple objects that are separate
from each other. Furthermore, the image contains only 2 colors. If this
is incorrect, please send me more details.
In either case, if you send me a sample image, I will try to suggest a good way to isolate each object.
You can either post the image here or send it to
support@LEADTOOLS.com
Note: If you post a file here, not all formats are accepted, so please put it in a ZIP file first.
Amin Dodin
Senior Support Engineer
LEAD Technologies, Inc.
#3
Posted
:
Thursday, February 2, 2006 10:48:44 PM(UTC)
Groups: Registered
Posts: 17
Amin,
I have separate objects (white color) and background (black color). I
want to get the objects (in a list or array or something else)
and some parameters for the objects (central positions, areas, edge
points...). Then, with these parameters, I will "clean" my image by
deleting some objects (example: small objects). Finally, I only have my
good objects and I can show them on the original image (not the
binarize one).
Regards
Julien
#4
Posted
:
Tuesday, February 7, 2006 6:31:34 AM(UTC)
Groups: Manager, Tech Support
Posts: 367
Was thanked: 1 time(s) in 1 post(s)
Julien,
Here's an idea to solve the problem:
Assuming you want objects that are totally separate, since identifying
2 connected objects as one is much easier than trying to identify them
as 2 separate objects.
If that's the case, one way to do this is to sweep through the image
pixels (e.g. from top left to bottom right) and do the following:
1. Using the Pixel property, determine if the pixel is white (i.e. part of a object).
2. If it's white, use SetRgnMagicWand method to create a region for that particular object.
3. Isolate the object as explained below, then fill the region with black to get it out of the way.
This way, every object is counted once and only once. For each object's
region, you can define a similar region in the original image and copy
the bitmap data from there to a secondary control for further
processing.
Note: To fill the region with black, simply call the Fill method, since it will affect the region alone
If you have a region in the image, calling some methods will affect the
region alone and leave the rest of the bitmap unchanged. Many functions
behave the same way. For a list of them, see the help topic "Raster
Images: Where the Region Preempts the Bitmap"
Amin Dodin
Senior Support Engineer
LEAD Technologies, Inc.
You cannot post new topics in this forum.
You cannot reply to topics in this forum.
You cannot delete your posts in this forum.
You cannot edit your posts in this forum.
You cannot create polls in this forum.
You cannot vote in polls in this forum.