Page 170 - Algorithm Collections for Digital Signal Processing Applications using MATLAB

4. Selected Applications 159

6.2.1 List of low level features computed in every sub blocks
of the image

1. Variance of the histogram of the hue content of the particular subblock is
computed.

2. Measurement of uniformity

p(z i) is the probability of gray value z i in the particular sub block of the
image . The uniformity of sub block of the image is measured using the
formula given below.

3. Measurement of Average Entropy

Average entropy is computed as described below.

4. Measurement of Relative smoothness

where, σ(z ) is the standard deviation of the gray values z in the
particular sub block of the image.

5, 6, 7, 8. Variance of the approximation Wavelet co-efficient

The image is subjected to discrete wavelet transformation (DWT) to
obtain approximation co-efficient (app), horizontal detail co-efficient (hor),
vertical detail co-efficient (ver) and diagonal (both horizontal and vertical)
detail co-efficient(dia) The variance of the approximation co-efficient ‘app’
th
is computed and considered as 5 feature. Similarly the variance of the detail
th
th
th
co-efficient ‘hor’,’ver’ and ‘dia’ are considered as the 6 , 7 and 8 features
respectively.