how to detect center of an object in an image and then crop the original image? using original image and green outline region

Question

Dayangku Nur Faizah Pengiran Mohamad 2024-5-20

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/2120491-how-to-detect-center-of-an-object-in-an-image-and-then-crop-the-original-image-using-original-image

回答： Aastha 2024-9-25，6:52

Hello.

I need to detect an object (suspicious area in an image). Then I need to find its centroid and then crop it the image into 256x256 pixels using image centroid as the centre of bounding box.

img = imread('1_245_original.jpg');
% colorspace
jmg = rgb2gray(img);
jm = mat2gray(jmg(:,:,1));
jm = imcomplement(jm);
% thresh
bw = imbinarize(jm, graythresh(jm));
% filter noise
bw = imopen(bw, strel('disk', 5));
bw = imfill(bw, 'holes');
% label every target
[L,num] = bwlabel(bw);
stats = regionprops(L);
figure; imshow(img, []);
for i = 1 : num
    % get rect 
    recti = stats(i).BoundingBox;
    % get cen
    ceni = stats(i).Centroid;
    % crop image
    imi = imcrop(img, round(recti));
    ims{i} = imi;
    % rect and cen
    hold on; rectangle('Position', recti, 'EdgeColor', 'c', 'LineWidth', 2);
    plot(ceni(1), ceni(2), 'yp', 'MarkerFaceColor', 'y', 'MarkerSize', 20);
end

I have tried out this code, but the cropping area was not in 256x256 pixels. and I want it to save at folder after detect the bounding box. Where can I put that code? Thanks.

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

Image Analyst 2024-5-20

You've asked this before. Finding dark regions is not a problem, nor is cropping them and resizing them. The problem is identifying which of the many dark regions in the image is the specific one you want, and you haven't specified what it is about that dark blob that distinguishes it from all the other dark blobs. Yes, I know you've said that it is diseased tissue but how does it LOOK different than all the others?

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Aastha 2024-9-25，6:52

1
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/2120491-how-to-detect-center-of-an-object-in-an-image-and-then-crop-the-original-image-using-original-image#answer_1522230

在 MATLAB Online 中打开

Hi @Dayangku Nur Faizah Pengiran Mohamad,

As I understand, for each proposed region you want to create a bounding box of size 256x256 pixels around the centroid and save it to disk. This can be accomplished by making the following additions to the given code:

For each region of proposals, the variable “recti” provides a tuple of the form (x, y, w, h), where (x, y) are the coordinates of the top-left corner, and (w, h) represent the width and height of the bounding box, respectively. To adjust the bounding box to a size of 256x256, you can make both the width and height equal to 256. This can be achieved by setting the width and height to 256 and adjusting the top-left corner (x, y) accordingly to keep the box centred around the centroid. You may refer to the code snippet below to do this:

ceni=stats(i).Centroid; 
topx=max(0,ceni(1)-128); % account for out of bounds index 
topy=max(0,ceni(2)-128); % account for out of bounds index 
recti=[topx topy 256 256]; 
imi = imcrop(img, round(recti)); 