Contour Detector (deprecated)

Detects nearly convex structures from a 'probability image' (where pixels of high intensity are likely to be part of part contour). (Deprecation: Will be replaced in future release!).


The maximum radius the objects to be detected can have.
Has impact on the precision of the contours to be detected. But more angles increase the runtime.
Line variance
The value one is usually appropriate.
Determines how much the detected objects are allowed to overlap. If two objects overlap more the specified, only the one with the higher score will be kept.
Minimal score
The minimal allowed score (between 0 and 1). All objects below that score will be disregarded.
Minimal area
Objects with a smaller area will be disregarded as well.
Distance equidistant seeding points
The distance between the seeding points over the image. The lower the distance, the more seeding points will be used causing more calculations. A good choice is the half size of the objects to be detected.
Smooth contour
Smoothes the contours by removing little perturbations.

Input Ports

Source Image and Seeds

Output Ports

Detected Segments


This node has no views


  • No workflows found



You want to see the source code for this node? Click the following button and we’ll use our super-powers to find it for you.