Represent a cross correlation as a mesh using imagesc() and mesh()

Question

Aaron Elliott 2020-11-1

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/632869-represent-a-cross-correlation-as-a-mesh-using-imagesc-and-mesh

Hi all,

I am doing template matching and trying to represent the cross correlaiton as a mesh using mesh() and imagesc(). The current code I have displays a mesh for a breif second but I don't think its correct. Any help/pointers would be much appreciated!

function BB = templateMatching(I,template)
    I = im2gray(I);
    template = im2gray(template);
    [Ix, Iy] = size(I);
    [Tx, Ty] = size(template);
    
    map = normxcorr2(template, I)
    map = imcrop(map,[Ty Tx Iy Ix]);
    mesh(map) 
    
    max_num = max(max(map));
    [r, c, v] = find(map==max_num);
    BB = insertShape(I, 'rectangle', [c r Ty Tx], 'LineWidth', 3);
end
% task file is below
clear all
close all
letters = imread('letters.JPG');
k = imread('K.JPG');
im1 = imread('image1.jpg');
im2 = imread('image2.jpg');
match1 = templateMatching(letters, k);
match2 = templateMatching(im1, im2);
figure(1)
subplot(1,3,1), imshow(letters), title('Image')
subplot(1,3,2), imshow(k), title('Template')
subplot(1,3,3), imshow(match1), title('BB')
figure(2)
subplot(1,3,1), imshow(im1), title('Image')
subplot(1,3,2), imshow(im2), title('Template')
subplot(1,3,3), imshow(match2), title('BB')