I could suss it out the old-fashioned math way, taking measurements from Photoshop's Info Palette, but that's too much like the torture we all had to go through in school. Depending on how much accuracy you need, you could always just draw an elliptical marquee that matches as closely as possible the curve you're curious about, and work it out from the Info Palette reading as well.
A radius is the point from the center of a shape to the edge of the shape. I'd suggest using a ruler (or ruler tool).
I hate Geometry. I'm an artist; not a mathemetician.
Thanks for the tip on another program to try. I will see if I can find it. Any suggestions where a poor college researcher could find a copy?
I have spent some time fighting with the ellipse tool, but I need to measure several hundred of these things in all, and trying to get the cicles to overlap is, like you said, far too reminiscent of class-time torture.
thanks again,
-kes
BLUDVLZ <nug...@aldebaran.net> wrote in message news:<ef9b...@WebX.aaQKaf7vad4>...
I am analysing some images for a biomedical research project. I would like to measure the radius of curvature of several dozen lines in each image.
In general, the radius of curvature varies continuously along a curve. Photoshop is definitely NOT the best program to use for this. You can manually measure the radius of curvature by drawing lines that are perpendicular to the curve at many points along the curve. They will intersect approximately at the local centers of the radius of curvature (the more lines you draw the better the approximations), and you can simply measure these local radii of curvature from those intersection points.
If you have a lot of these measurements to do, you do need a specialized program as a labor saver, and for more accurate results. Perhaps the HotDoor CADTools Illustrator plugin would be capable -- I don' know. There are other programs you could consider. For example, Mathematica is a major tool, and MathCAD is a lesser tool with a lesser price.
One mathematical approach you could follow would be to curve fit an analytical form to points sampled on your curves and derive an analytical form for your radius of curvature from that analytical form. There are a lot of numerical methods you could use. None of which involve Photoshop.
-- Burton --