红外小目标探测经典方法LCM(Local Contrast Method)的MATLAB代码

% 经典论文"A Local Contrast Method for Small Infrared Target Detection"中的Algorithm 2 % 输入patch的尺寸必须是3的倍数，考虑论文定义u的大小只能是3x3,5x5,7x7,9x9，patch的尺寸对应9x9,15x15,21x21,27x27 function [C_hat,max_margin] = MLCM_computation(I_MLCM_in) I_MLCM_in = double(I_MLCM_in); [row,col] = size(I_MLCM_in); scales = [9,15,21,27]; % patch的尺寸有9x9,15x15,21x21,27x27 l_max = size(scales); % 对应论文lmax % Compute Cl according to Algorithm 1 C_map_scales = zeros(row,col,l_max(2)); for i = 1:l_max(2) % 对应不同尺度 for j = 1:row-scales(i)+1 % 单一尺度下以patch为单位做遍历 for k = 1:col-scales(i)+1 temp_patch = I_MLCM_in(j:j+scales(i)-1, k:k+scales(i)-1); temp_patch_1 = LCM_computation(temp_patch); % 对patch执行Algorithm 1 C_map_scales(j:j+scales(i)-1, k:k+scales(i)-1,i) = temp_patch_1; end end end % 这部分计算，生成4张对比度图，其中尺度最大的对比度图有效像元数最小，每个方向减去(scales(4)-1)/2=13 max_margin = (scales(4)-1)/2; % 输出4种尺度的对比图 figure() [X,Y] = meshgrid(1:1:row,1:1:col); subplot(2,2,1); mesh(X,Y,C_map_scales(:,:,1)); axis([0 row 0 col 0 255]);xlabel('row');ylabel('col');zlabel('value'); title('v=3x3 Contrast Map'); subplot(2,2,2); mesh(X,Y,C_map_scales(:,:,2)); axis([0 row 0 col 0 255]);xlabel('row');ylabel('col');zlabel('value'); title('v=5x5 Contrast Map'); subplot(2,2,3); mesh(X,Y,C_map_scales(:,:,3)); axis([0 row 0 col 0 255]);xlabel('row');ylabel('col');zlabel('value'); title('v=7x7 Contrast Map'); subplot(2,2,4); mesh(X,Y,C_map_scales(:,:,4)); axis([0 row 0 col 0 255]);xlabel('row');ylabel('col');zlabel('value'); title('v=9x9 Contrast Map'); % 对4种尺度对比图的共同部分取最大值，作为输出 C_hat = zeros(row-scales(4)+1,col-scales(4)+1); for i = 1:row-scales(4)+1 for j = 1:row-scales(4)+1 temp = [C_map_scales(i+max_margin,j+max_margin,1);... C_map_scales(i+max_margin,j+max_margin,2);... C_map_scales(i+max_margin,j+max_margin,3);... C_map_scales(i+max_margin,j+max_margin,4)]; C_hat(i,j) = max(temp); end end end