2.3.1. Normalisation
The intensity of the digital x-ray image, acquired from various
subjects, generally varied; the grey level varied over large distances
as a result of radiological artefacts and the fat tissue
projection on the radiograph corresponded to the low frequency
noise of the image. Hence, the technique proposed by Geraets and
colleagues was used to eliminate the low frequency noise of the
image when the trabecular component of the image was taken
into account [14]. Then the intensity normalisation, incorporating
the histogram specification technique detailed by Debashis
and colleagues [15], was used in order to obtain the desired
histogram by maximising a measure that represented increase in
information (trabecular structure enhancement) and decrease in
ambiguity, which are contradictory indicators of contrast enhancement
(histogram equalisation). This approach was independently
applied to each individual 33 sub-block in order to obtain
normalised ROI (IN).