Thresholding of Medical Images - Medical Image Processing: Advanced Fuzzy Set Theoretic Techniques - page 128

For all threshold grey levels, intuitionistic fuzzy divergence is calculated.
The threshold level corresponding to the minimum divergence is selected as
the optimal threshold.
This is calculated for all the windows. The final threshold is the optimal
threshold minus (1/4)th of the standard deviation of the window:

1
4

− ⎛

⎞
⎟ ×

opt

Final_Th Th = S.D.

⎜

window

window

where S.D. is the standard deviation
Had the fraction not been there, the images will be more segmented, that
is, all the blood vessels are not extracted. Thus, the fraction of 1/4 has been
used to obtain a good segmented image.

Example 6.5

Two examples of blood vessel and blood cell images are shown in
Figure 6.5 to illustrate the efficacy of window-based thresholding
using intuitionistic fuzzy set on medical images. It is observed that the
blood vessels/cells are clearly thresholded on poor contrasted medical
images.

(a) (b)

FIGURE 6.5
(a) Blood vessel image and (b) intuitionistic fuzzy local thresholding method by Chaira.