Object description - Feature Extraction and Image Processing - page 281

%Region descriptors (compactness)

function RegionDescrp(inputimage)

%Image size
[rows,columns]=size(inputimage);

%area
A=0;
for x=1:columns
for y=1:rows
if inputimage(y,x)==0 A=A+1; end

end
end

%Obtain Contour
C=Contour(inputimage);

%Perimeter & mean
X=C(1,:); Y=C(2,:); m=size(X,2);

mx=X(1); my=Y(1);
P=sqrt((X(1)-X(m))^2+(Y(1)-Y(m))^2);
for i=2:m
P=P+sqrt((X(i)-X(i-1))^2+(Y(i)-Y(i-1))^2);
mx=mx+X(i); my=my+Y(i);
end

mx=mx/m; my=my/m;

%Compactness
Cp=4*pi*A/P^2;

%Dispersion
max=0; min=99999;

for i=1:m
d=((X(i)-mx)^2+(Y(i)-my)^2);
if (d>max) max=d; end
if (d<min) min=d; end
end

I=pi*max/A;
IR=sqrt(max/min);

%Results
disp('perimeter='); disp(P);

disp('area='); disp(A);

disp('Compactness='); disp(Cp);

disp('Dispersion='); disp(I);

disp('DispersionR='); disp(IR);

Code 7.4 Evaluating basic region descriptors