This method performs an adaptive gray value segmentation. This procedure is particularly well suited to the segmentation of small structures against a varying background.

The function initially applies a low pass filter and then subtracts this low-pass-filtered image from the input image. The effect of this function mainly depends on the size of the filter matrix: Select a low value for Kernel Size to segment small regions or regions with low gray value contrast from the background. Select a higher value for Kernel Size to segment larger regions from the background.

Parameter

Description

Kernel Size

Matrix size of the low pass filter in x- and y-direction symmetrically around the pixel in question. Determines the extent of the smoothing effect. As the affected pixel is at the center, the edge length of the filter matrix is always an odd number. If an even number is entered via the keyboard, the value is always set to the next highest odd number.

Threshold

This value defines the gray value difference between the regions to be detected and the background. Segmented pixels are set to the maximum gray value (white), whilst other pixels are set to the gray value 0.

The following parameters are only visible if the Show All mode is activated:

Parameter

Description

Create Binary

Activated: The resulting image is a binary image. Pixels within the calculated gray level range are set to the maximum gray value (white), whilst pixels outside it are set to the gray value 0.

Deactivated: The resulting image is of the same type as the input image. Pixels within the calculated gray level range are set to the original gray value. Pixels outside it are set to 0.

Invert Result

Activated: Inverts the effect of the function. The segmented regions will be given the value 0, and all other pixels the gray value white or the gray value/color of the input image.