Module: Regional Maxima ()

Description:

This module computes the regional or relative maxima in a grayscale image and creates a binary image containing these maxima.

A regional maximum C is a set of connected pixels such that:

1. Pixels belonging to C have the same intensity .
2. Pixels connected to C, but not belonging to C (neighbors), have an intensity lower than .

Figure 1: Examples of regional maxima in 1 and 2D.

This algorithm is based on [1] and uses a recursive method combined with a geodesic propagation. To avoid getting too many regions in the output image, input should first be smoothed with a low-pass filter or with the dual numerical reconstruction algorithm.

[1] B. Laÿ. Recursive Algorithms in Mathematical Morphology. In Acta Stereologica Vol. 6/III, pages 691--696, Caen, France, Sept. 1987. 7th International Congress For Stereology.

See also: Regional Minima, Reconstruction By Dilation.

Connections:

Input Image [required]
The image to be processed. Supported types are: grayscale/color image (Uniform Scalar Field/Uniform Color Field), binary (Uniform Label Field with 2 labels) and label image (Uniform Label Field).

Ports:

Interpretation

This port specifies whether the input will be interpreted as a 3D volume or a stack of 2D images for processing.

• "3D": the module configuration is set to 3D. The image will be processed as a whole in 3D.
• "XY planes": the module configuration is set to 2D. The image will be processed slice per slice.

Neighborhood

In 3D configuration, this port refers to the type of connectivity considered for processing adjacent voxels:

• 6: voxels with a common face are considered connected
• 18: voxels with at least one common edge are considered connected
• 26: voxels with at least one common vertex are considered connected