2.3.2. Trabecular bone segmentation
Once the normalisation was over, the local contrast enhancement
transformation (LCET) technique was used to enhance both
the bone contrast attribute and total image intensity. The aim was
to define the point transformation dependent on window (W),
such that the distribution was localised around the mean of the
intensity and covered the entire intensity range. The latent
hypothesis is that W is large enough to contain a statistically
representative distribution of the local variation of grey levels, yet
small enough to be unaffected by the gradual change of contrast
between the centre and periphery of the image. If each pixel in the
normalised ROI (IN) is p, and small running window is W, a new
segmented image (IF) would result as follows