|
Pink 0.9
|
enhances linear structures in a grayscale image More...
enhances linear structures in a grayscale image
Usage: crestvol in.pgm connex radius out.pgm
Description: Enhances linear structures in a grayscale image. The algorithm is the following:
Let F be the original image ("in.pgm")
Let O be the final result ("out.pgm")
Let r be the value of parameter "radius"
Compute k iterations of homotopic thinning, result S
Lower all separating points p in S down to alpha8m(S, p), result Sp
For each point x of S which is separating
Let Dx be the image with D(x) = F(x), and D(y) = 0 for all y != x
Let Fx be the geodesic reconstruction of Dx under F
O(x) = volume((Fx-Sp).Br(x))
where Br(x) is the disc centered on x with radius r
EndFor
References:
"Topological operators for the detection of
curvilinar structures in grayscale images", in preparation.
Types supported: byte 2d
Category: topogray
1.7.3