X&Y location of intersecting lines

[TLowe]

I'm new to ImageJ but learning something new everyday, so bare with me. I have an application where I have an image(actually I'm soon going to have thousands of images) that will all basically look the same; they are two intersecting lines. I have a routine that I recorded that performs a function on the intersection of these lines. In the routine, I have the pixel address of the intersection hardcoded. Ideally, I'd like it to find the address of the intersection & run the routine on that address(ie because the location of this intersection will 'wander' a bit from image to image, due to positioning of the camera). Has anyone written a plugin/script/macro for such a thing?
After reading all the various applications, I thought I'd ask.
Thanks
Tom

[Gabriel Landini]

On Tuesday 26 Aug 2014 16:53:15 you wrote:

> I'm new to ImageJ but learning something new everyday, so bare with me. I
> have an application where I have an image(actually I'm soon going to have
> thousands of images) that will all basically look the same; they are two
> intersecting lines. I have a routine that I recorded that performs a
> function on the intersection of these lines. In the routine, I have the
> pixel address of the intersection hardcoded. Ideally, I'd like it to find
> the address of the intersection & run the routine on that address(ie because
> the location of this intersection will 'wander' a bit from image to image,
> due to positioning of the camera). Has anyone written a plugin/script/macro
> for such a thing?

Can you post an example image, please?
Cheers

Gabriel

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[TLowe]

[Gabriel Landini]

On Wednesday 27 Aug 2014 04:41:58 TLowe wrote:
> <http://imagej.1557.x6.nabble.com/file/n5009376/Untitled.jpg>

This seems to find the intersectionpn that image:

run("8-bit");
run("Invert");
run("Minimum...", "radius=4");
run("Gaussian Blur...", "sigma=6");
run("Find Maxima...", "noise=10 output=[Point Selection]");
getSelectionCoordinates(x,y);
print(x[0],y[0]);

You might have to adjust some of parameters above (radius and sigma) depending
if the line thickness is always the same.
Perhaps converting to 32-bit instead of 8-bit is better, not sure.

Hope it helps.
Cheers

Gabriel

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html