HOW TO CALCULATE THE DICE SIMILARITY OF THE IMAGES SUBPLOT.

Hi all, I have 2 data set logical images(binary images). EACH DATA SET HAVE 23 IMAGES. I want to check the dice similarity.

Below is the code for aorigional images.

%% first, read the image data and labelled images
clc
clear all
dataSetDir = fullfile('C:\Users\Akmal\Desktop\I-131 256 28.02.2020\I-131 SPECT NEMA VALIDATION 01112019 256X256 26.09.2021 petang');
imageDir = fullfile(dataSetDir,'labelledimages');
imds = imageDatastore(imageDir);
% view data set images origional
figure
for i = 1:23
    subplot(5,5,i)
    I = readimage(imds,i);
    imshow(I)
    title('training labels')
end

The the second one images code is below

%% second, read the binary images after segmentation
dataSetDir1 = fullfile('C:\Users\Akmal\Desktop\I-131 256 28.02.2020\I-131 SPECT NEMA VALIDATION 01112019 256X256 26.09.2021 petang');
imageDir1 = fullfile(dataSetDir1,'bnwaftersegmentation');
imds1 = imageDatastore(imageDir1);
% view data set images origional
figure
for ii = 1:23
    subplot(5,5,ii)
    II = readimage(imds1,ii);
    imshow(II)
    title('binary labels')
end

Then i used code below to know the dice similarity, but the answer is 0

similarity = dice(I, II)
similarity =
0

But I try test just one image (let say image number 11), its work.

s = imread('11.png');
d = imread('11.png');
similarity = dice(s,d)
similarity =
0.15119

ANYONE CAN HELP ME HOW TO CALCULATE THE TOTAL DICE SIMILSRITY FOR ALL 23 IMAGES

Answer 1

[yanqi liu]

sir，please check the follow code to get some information

for more information,you can read the book 《计算机视觉与深度学习实战》，Thank you！

%% first, read the image data and labelled images
clc
clear all
dataSetDir = fullfile('C:\Users\Akmal\Desktop\I-131 256 28.02.2020\I-131 SPECT NEMA VALIDATION 01112019 256X256 26.09.2021 petang');
imageDir = fullfile(dataSetDir,'labelledimages');
imds = imageDatastore(imageDir);
% view data set images origional
Is = [];
figure
for i = 1:23
    subplot(5,5,i)
    I = readimage(imds,i);
    Is{end+1} = I;
    imshow(I)
    title('training labels')
end

%% second, read the binary images after segmentation
dataSetDir1 = fullfile('C:\Users\Akmal\Desktop\I-131 256 28.02.2020\I-131 SPECT NEMA VALIDATION 01112019 256X256 26.09.2021 petang');
imageDir1 = fullfile(dataSetDir1,'bnwaftersegmentation');
imds1 = imageDatastore(imageDir1);
% view data set images origional
Is2 = [];
figure
for ii = 1:23
    subplot(5,5,ii)
    II = readimage(imds1,ii);
    Is2{end+1} = II;
    imshow(II)
    title('binary labels')
end

%% compare the dice similarity for every slice, like 1 with 1, 2 with 2, 3 with 3....and so on till 23 with 23..
similarity = [];
for i = 1 : 23
    similarity(i) = dice(Is{i}, Is2{i});
    fprintf('the dice similarity for %d with %d is %.3f\n', i, i, similarity(i));
end

Answer 2

[Image Analyst]

Just read in I before the loop, then put the line

similarity(ii) = dice(I, II)

inside the loop.

[mohd akmal masud]

%% first, read the image data and labelled images
clc
clear all
dataSetDir = fullfile('C:\Users\Akmal\Desktop\I-131 256 28.02.2020\I-131 SPECT NEMA VALIDATION 01112019 256X256 26.09.2021 petang');
imageDir = fullfile(dataSetDir,'labelledimages');

imds = imageDatastore(imageDir);
% view data set images origional
I = readimage(imds);

You mean like this? but got error

Not enough input arguments.

Error in matlab.io.datastore.ImageDatastore/readimage (line 32)
[data, info] = matlab.io.datastore.FileBasedDatastore.errorHandlerRoutine(imds,ME,index);

[Image Analyst]

Try this:

imds = imageDatastore(imageDir)
numFiles = length(imds.Files);
fprintf('Found %d image files.\n', numFiles);
% Read in first image.
thisFullFileName = imds.Files{1};
fprintf('Reading image #%d of %d : "%s" ...\n', ...
    1, numFiles, thisFullFileName);
firstImage = imread(thisFullFileName);
imshow(firstImage);

% Read in and view subsequent images.
for k = 2 : numFiles
    % Get the file name.
    thisFullFileName = imds.Files{k};
    fprintf('Reading image #%d of %d : "%s" ...\n', ...
        k, numFiles, thisFullFileName);
    thisImage = imread(thisFullFileName);
    imshow(thisImage);
    % similarity(k) = dice(firstImage, thisImage);
end

[mohd akmal masud]

sir, What I mean is, I want compare all the image for set images

this one is expected (we known the area of the slice), have 23 images

%% first, read the image data and labelled images
clc
clear all
dataSetDir = fullfile('C:\Users\Akmal\Desktop\I-131 256 28.02.2020\I-131 SPECT NEMA VALIDATION 01112019 256X256 26.09.2021 petang');
imageDir = fullfile(dataSetDir,'labelledimages');
imds = imageDatastore(imageDir);
% view data set images origional
figure
for i = 1:23
    subplot(5,5,i)
    I = readimage(imds,i);
    imshow(I)
    title('training labels')
end

this one is what I get after did some segmentation technique.

%% second, read the binary images after segmentation
dataSetDir1 = fullfile('C:\Users\Akmal\Desktop\I-131 256 28.02.2020\I-131 SPECT NEMA VALIDATION 01112019 256X256 26.09.2021 petang');
imageDir1 = fullfile(dataSetDir1,'bnwaftersegmentation');
imds1 = imageDatastore(imageDir1);
% view data set images origional
figure
for ii = 1:23
    subplot(5,5,ii)
    II = readimage(imds1,ii);
    imshow(II)
    title('binary labels')
end

Then I want compare the dice similarity for every slice, like 1 with 1, 2 with 2, 3 with 3....and so on till 23 with 23..

I used this code for combine all the 23 images similarity, but the answer is 0.

similarity = dice(I, II)
similarity =
0

But when i test just one pair of image only, (let say image number 11 with 11), its can show the similarity coefficient.

I11 = imread('11.png'); %read from folder expected labelled
II11 = imread('11.png'); %read from folder after segmentation
similarity = dice(I11,II11)
similarity =
0.15119

can you got my point sir?