% gets property vector for a binary shape in an image
function vec = getproperties(Image)

     [H,W] = size(Image);
     area = bwarea(Image);
     perim = bwarea(bwperim(Image,4));
     
     % compactness
     compactness = perim*perim/(4*pi*area);

     % rescale properties so all have size proportional
     % to image size
%     vec = [4*sqrt(area), perim, H*compactness];

% get scale-normalized complex central moments
c11 = complexmoment(Image,1,1) / (area^2);
c20 = complexmoment(Image,2,0) / (area^2);
c30 = complexmoment(Image,3,0) / (area^2.5);
c21 = complexmoment(Image,2,1) / (area^2.5);
c12 = complexmoment(Image,1,2) / (area^2.5);
%c=[c11,c20,c30,c21,c12]

  % get invariants, scaled to [-1,1] range
ci1 = real(c11);
ci2 = real(1000*c21*c12);
tmp = c20*c12*c12;
ci3 = 10000*real(tmp);
ci4 = 10000*imag(tmp);
tmp = c30*c12*c12*c12;
ci5 = 1000000*real(tmp);
ci6 = 1000000*imag(tmp);

%ci=[ci1,ci2,ci3,ci4,ci5,ci6]

%     vec = [compactness,i1,i2,i3,i4,i5,i6,i7];
     vec = [compactness,ci1,ci2];         %only use 3 as only have 4 samples