✅作者簡介：熱愛科研的Matlab仿真開發者，修心和技術同步精進，matlab項目合作可私信。
🍎個人首頁：Matlab科研工作室
🍊個人信條：格物緻知。
⛄ 内容介紹
針對傳統Kapur熵和oust在多門檻值圖像分割算法中存在運算量大、計算效率低以及精度不高等問題,提出了一種基于電磁算法的多級門檻值圖像分割方法,該方法采用Kapur熵作為計算适應度的目标函數,通過引入電磁算法求解目标函數最大化時的全局最優問題.實驗結果表明:相對于其他方法,本文方法在多個評價名額上都有很好的性能展現,并且本文方法在保證較好分割效果的同時,計算效率明顯提升.
⛄ 部分代碼
%Diego Oliva, Erik Cuevas, Gonzalo Pajares, Daniel Zaldivar, Valent韓 Osuna.
%A Multilevel Thresholding algorithm using electromagnetism optimization
%Universidad Complutense de Madrid / Universidad de Guadalajara
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%The algorithm was published as:
%Diego Oliva, Erik Cuevas, Gonzalo Pajares, Daniel Zaldivar, Valent韓 Osuna.
%A Multilevel Thresholding algorithm using electromagnetism optimization,
%Journal of Neurocomputing, 139, (2014), 357-381.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
function [fitR, fitBestR, fitG, fitBestG, fitB, fitBestB] = fitnessIMG(I, N, Lmax, level, xR, probR, xG, probG, xB, probB)
%Metodo de Otsu
%Evalua poblaciones xR, xG, xB, en la funcion objetivo para obtener
%fitR, fitG, fitB, dependiendo si la imagen es RGB o escala de grises
for j = 1:N
if size(I,3) == 1
%grayscale image
fitR(j) = sum(probR(1:xR(j,1))) * (sum((1:xR(j,1)) .* probR(1:xR(j,1)) / sum(probR(1:xR(j,1)))) - sum((1:Lmax) .* probR(1:Lmax)) ) ^ 2;
for jlevel = 2:level - 1
fitR(j) = fitR(j) + sum(probR(xR(j,jlevel - 1) + 1:xR(j,jlevel))) * (sum((xR(j,jlevel - 1) + 1:xR(j,jlevel)) .* probR(xR(j,jlevel - 1) + 1:xR(j,jlevel)) / sum(probR(xR(j,jlevel - 1) + 1:xR(j,jlevel)))) - sum((1:Lmax) .* probR(1:Lmax))) ^ 2;
end
fitR(j) = fitR(j) + sum(probR(xR(j,level-1) + 1:Lmax)) * (sum((xR(j,level - 1) + 1:Lmax) .* probR(xR(j,level - 1) + 1:Lmax) / sum(probR(xR(j,level - 1) + 1:Lmax))) - sum((1:Lmax) .* probR(1:Lmax))) ^ 2;
% if isnan(fitR(j))
% fitR(j)=eps;
% end
fitBestR(j) = fitR(j);
elseif size(I,3) == 3
%RGB image
fitR(j) = sum(probR(1:xR(j,1))) * (sum((1:xR(j,1)) .* probR(1:xR(j,1)) / sum(probR(1:xR(j,1)))) - sum((1:Lmax) .* probR(1:Lmax))) ^ 2;
for jlevel = 2:level - 1
fitR(j) = fitR(j) + sum(probR(xR(j,jlevel - 1) + 1:xR(j,jlevel))) * (sum((xR(j,jlevel - 1) + 1:xR(j,jlevel)) .* probR(xR(j,jlevel - 1) + 1:xR(j,jlevel)) / sum(probR(xR(j,jlevel - 1) + 1:xR(j,jlevel)))) - sum((1:Lmax) .* probR(1:Lmax))) ^ 2;
end
fitR(j) = fitR(j) + sum(probR(xR(j,level-1) + 1:Lmax)) * (sum((xR(j,level - 1) + 1:Lmax) .* probR(xR(j,level - 1) + 1:Lmax) / sum(probR(xR(j,level - 1) + 1:Lmax))) - sum((1:Lmax) .* probR(1:Lmax))) ^ 2;
fitBestR(j) = fitR(j);
fitG(j) = sum(probG(1:xG(j,1))) * (sum((1:xG(j,1)) .* probG(1:xG(j,1)) / sum(probG(1:xG(j,1)))) - sum((1:Lmax) .* probG(1:Lmax))) ^ 2;
for jlevel = 2:level - 1
fitG(j) = fitG(j) + sum(probG(xG(j,jlevel - 1) + 1:xG(j,jlevel))) * (sum((xG(j,jlevel - 1) + 1:xG(j,jlevel)) .* probG(xG(j,jlevel - 1) + 1:xG(j,jlevel)) / sum(probG(xG(j,jlevel - 1) + 1:xG(j,jlevel)))) - sum((1:Lmax) .* probG(1:Lmax))) ^ 2;
end
fitG(j) = fitG(j) + sum(probG(xG(j,level - 1) + 1:Lmax)) * (sum((xG(j,level-1) + 1:Lmax) .* probG(xG(j,level - 1) + 1:Lmax) / sum(probG(xG(j,level - 1) + 1:Lmax))) - sum((1:Lmax) .* probG(1:Lmax))) ^ 2;
fitBestG(j) = fitG(j);
fitB(j) = sum(probB(1:xB(j,1))) * (sum((1:xB(j,1)) .* probB(1:xB(j,1)) / sum(probB(1:xB(j,1)))) - sum((1:Lmax) .* probB(1:Lmax))) ^ 2;
for jlevel = 2:level - 1
fitB(j) = fitB(j) + sum(probB(xB(j,jlevel - 1) + 1:xB(j,jlevel))) * (sum((xB(j,jlevel - 1) + 1:xB(j,jlevel)) .* probB(xB(j,jlevel - 1) + 1:xB(j,jlevel)) / sum(probB(xB(j,jlevel - 1) + 1:xB(j,jlevel)))) - sum((1:Lmax) .* probB(1:Lmax))) ^ 2;
end
fitB(j) = fitB(j) + sum(probB(xB(j,level - 1) + 1:Lmax)) * (sum((xB(j,level - 1) + 1:Lmax) .* probB(xB(j,level - 1) + 1:Lmax) / sum(probB(xB(j,level - 1) + 1:Lmax))) - sum((1:Lmax) .* probB(1:Lmax))) ^ 2;
fitBestB(j) = fitB(j);
end
end
if size(I,3) == 1
%Imagen escala de Grises
fitR = fitR';
fitBestR = fitBestR';
elseif size(I,3) == 3
% Imagen RGB
fitR = fitR';
fitBestR = fitBestR';
fitG = fitG';
fitBestG = fitBestG';
fitB = fitB';
fitBestB = fitBestB';
end
⛄ 運作結果
