Page 167 - Algorithm Collections for Digital Signal Processing Applications using MATLAB

156 Chapter 4

6. SEARCH ENGINE USING INTERACTIVE
GENETIC ALGORITHM

Consider the small Image database stored with Low-level feature vectors
associated with every image. Low-level feature vector associated with every
image is the 1D vector whose elements are the statistical measurements like
variance, kurtosis, and higher order moments etc., computed from the sub
blocks of the image. Low-level feature vector associated with every image in
the database acts as the index for searching the particular image from the
database. Interactive Genetic algorithm based searching procedure is
described below.

6.1 Procedure

1. The front panel of the search engine consists of finite number of images
displayed. (Say 8 images) as shown in the figure given below.
2. The user is asked to assign the rank for the individual images. The value
ranges from 0 to100 based on the user’s interest. (i.e.) User assigns higher

Figure 4-14. Sample front panel