有一个图像分块的代码,可以直接将一幅图像分为5*5的小块,代码如下: %[FileName,PathName] = uigetfile('*.*','Select the image'); Im=imread('Bicubic flowers.bmp'); imshow(Im) hold on L = size(Im); height=5; width=5; max_row = floor(L(1)/height);%实验图片为800*1280,则max_row=5,max_col=8 max_
1. block_divide % 返回的块向量构成的矩阵,其维度信息为 K^2 * N,每一列由块构成的列向量 function P = block_divide(I, K) r = size(I, 1)/K; c = size(I, 2)/K; P = zeros(K^2, r*c); f = @(A) A(:); for i = 1:r, for j = 1:c, P(:, (i-1)*c+j) = f(I(K*(i-1)+1:K*i, K*(j-1)+1:K*j)); end end e
一 前言 最近收到审稿人的修改意见,其中一条为<RC: There were only five images evaluated in the experiment, and I recommend increasing to twenty to further evaluate the algorithm performance.>他说我论文只有五副图像用来评价算法性能,推荐我至少用20副图像来仿真算法效果. 但绝大多数SCI论文的实验部分都是五副红外图像,三个参考算法,即每一张红外图像对