Leukocyte Detection Through an Evolutionary Method - Complex System Modelling and Control Through Intelligent Soft Computations - page 148

Step 1: Set the DE parameters F = 0.25 and CR = 0.8

Step 2:

Initialize the population of m individuals E k

where each
decision variable p 1 , p 2 , p 3 , p 4 and p 5 of E a is set randomly within the interval
[1, N p ]. All values must be integers. Considering that k = 0 and a 2ð 1 ; 2 ; ...; m Þ

¼f E 1 ; E 2 ; ...; E k m g

Step 3:

Evaluate the objective value J(E a ) for all m individuals, and determining the E best ; k
showing the best fitness value, such that
E best ; k E k J E best ; k E 1 Þ; E 2 Þ; ...; E k m Þ 2 ð Þ¼ min J ð J ð J ð

Step 4:

Generate the trial population T = {, , , }

m

TT T :

12

for ( i =1; i < m +1; i ++)
do r =floor(rand(0,1) ⋅

); while ( 1 ri = );

do r =floor(rand(0,1) ⋅ ); while (( 2 ri = ) or ( 2 rr = )); 2

jrand=floor( 5 ⋅

rand(0,1));

for ( j =1; j <6; j ++) // generate a trial vector
if (rand(0,1)<= CR or j =jrand)

TE FEE =

best k ,

+ ⋅ (

k

−

k

);

ji , j jr , jr ,

1 2

else

TE =

k

, ; ji , ji

end if
end for

end for

Step 5: Evaluate the fitness values J(Ti) i if ) of all trial individuals. Check

all individuals. If a candidate parameter set is not physically plausible, i.e. out of
the range [1, N p ], then an exaggerated cost function value is returned. This aims to
eliminate

2 f 1 ; 2 ; ...; m g

'' unstable ” individuals.

Step 6:

k + 1 k + 1 k + 1 k

m

+ 1

Select the next population E = {, EE E , , } :

1 2

for ( i =1; i< m +1; i++)
if (( ) JT JE < ( ))

k

i i

ET

k + 1

= else 1
k

i i

+ k

EE =

i i

end if
end for

If the iteration number (NI) is met, then the output E best,k is the solution (an actual
ellipse contained in the image), otherwise go back to Step 3

Step 7:

4 The White Blood Cell Detector

In this section, the complete WBC detection strategy is described. Such a strategy
combines a segmentation method with the ellipse detection approach presented in
Sect. 3 .