a=imread('a6.jpg'); m=ones(600,800,3); c=255.*m; e=ones(600,800,3) for i=1:600; for j=1:800; if a(i,j,1)<100 if a(i,j,1)>55 c(i,j,1)=a(i,j,1); if a(i,j,2)<104 if a(i,j,2)>75 c(i,j,2)=a(i,j,2); if a(i,j,3)<90 if a(i,j,3)>56 c(i,j,3)=a(i,j,3); end end end end end end end end e=c./255; colormap(jet); image(e); ...
You cannot get it from thresholding or even combined with texture classification. There is just too much higher level knowledge required. Maybe try this: http://www.mathworks.com/matlabcentral/fileexchange/37197-dem-diffused-expectation-maximisation-for-image-segmentation or similar. But even that won't be perfect. I suggest you use ginput or roipoly to manually locate them (unless you have thousands of images).
Play games and win prizes!Learn more