how to calculate the area binary if some images have different pattern

Hi all I have 4 images that all images have 4 area(like circle) as below.

Then, I used this coding below to get the area for all the 4 images that every images have 4 circle

%% TO GET THE VOLUME SEGMENTATION AFTER DEEP LEARNING PERFORM(blob all the sphere binary)
alldice=[]
acc=[]
Ts = [];
Data = [];

for ii=1:4
    subplot(1,4,ii)
    I = readimage(IMDS,ii);
    [C,scores] = semanticseg(I,net1);
    
    outt2=C=="foreground";
    st2=strel('disk',5);
    outt22=imerode(outt2,st2);
%       outt22(:,:,ii) = imerode(outt2,st2);

    title('output')
    imshow(outt22)

info = regionprops(outt22,'Boundingbox') ;
imshow(outt22)
hold on

allbb=[]
for k = 1 : length(info)
     BB = info(k).BoundingBox;
     rectangle('Position', [BB(1),BB(2),BB(3),BB(4)],'EdgeColor','r','LineWidth',2) ;
     allbb=[allbb; BB]
end

	vol = [];
	for i=1:4
    
		zz=imcrop(outt22,allbb(i,:));
		
		stats = regionprops('table',zz,'Centroid',...
			'MajorAxisLength','MinorAxisLength');
		T = regionprops('table', outt22,'Area','Centroid')
		m1=stats.MajorAxisLength;
		m2=stats.MinorAxisLength;
		v1=stats.MinorAxisLength;
		
		vol(i)=m1*m2*(v1/2);					
    
	end
	ss=sprintf('the volume %d is %6f,%d is %6f,%d is %6f,%d is %6f',1,vol(1),2,vol(2),3,vol(3),4,vol(4));
		
	msgbox(ss);
end

Its work. And the answer like below

T =

4×2 table

Area        Centroid    
____    ________________

15     105.73    132.53
21     126.52    145.38
64     147.12    108.86
31     147.42     133.9


T =

4×2 table

Area        Centroid    
____    ________________

15     105.73    132.53
21     126.52    145.38
64     147.12    108.86
31     147.42     133.9


T =

4×2 table

Area        Centroid    
____    ________________

15     105.73    132.53
21     126.52    145.38
64     147.12    108.86
31     147.42     133.9


T =

4×2 table

Area        Centroid    
____    ________________

15     105.73    132.53
21     126.52    145.38
64     147.12    108.86
31     147.42     133.9

But when I tried the same code into this images below. It have error.

**NOTE: the first and second image just have 3 circle, the 2 others have 4 circle.

T =

3×2 table

Area        Centroid    
____    ________________

6     125.67    146.33
92     147.77    109.37
52     147.15    134.06


T =

3×2 table

Area        Centroid    
____    ________________

6     125.67    146.33
92     147.77    109.37
52     147.15    134.06


T =

3×2 table

Area        Centroid    
____    ________________

6     125.67    146.33
92     147.77    109.37
52     147.15    134.06

Index in position 1 exceeds array bounds (must not exceed 3).

anyone can help me?

Answer 1

[yanqi liu]

%% TO GET THE VOLUME SEGMENTATION AFTER DEEP LEARNING PERFORM(blob all the sphere binary)
alldice=[]
acc=[]
Ts = [];
Data = [];

for ii=1:4
    subplot(1,4,ii)
    I = readimage(IMDS,ii);
    [C,scores] = semanticseg(I,net1);
    
    outt2=C=="foreground";
    st2=strel('disk',5);
    outt22=imerode(outt2,st2);
    %       outt22(:,:,ii) = imerode(outt2,st2);
    
    title('output')
    imshow(outt22)
    
    info = regionprops(outt22,'Boundingbox') ;
    imshow(outt22)
    hold on
    
    allbb=[]
    for k = 1 : length(info)
        BB = info(k).BoundingBox;
        rectangle('Position', [BB(1),BB(2),BB(3),BB(4)],'EdgeColor','r','LineWidth',2) ;
        allbb=[allbb; BB]
    end
    
    vol = [];
    for i=1:size(allbb,1)
        
        zz=imcrop(outt22,round(allbb(i,:)));
        
        stats = regionprops('table',zz,'Centroid',...
            'MajorAxisLength','MinorAxisLength');
        T = regionprops('table', outt22,'Area','Centroid')
        m1=stats.MajorAxisLength;
        m2=stats.MinorAxisLength;
        v1=stats.MinorAxisLength;
        
        vol(i)=m1*m2*(v1/2);
        
    end
    ss=sprintf('the volume %d is %6f,%d is %6f,%d is %6f,%d is %6f',1,vol(1),2,vol(2),3,vol(3),4,vol(4));
    
    msgbox(ss);
end

[mohd akmal masud]

sir, got error

Index exceeds the number of array elements (2).

[yanqi liu]

%% TO GET THE VOLUME SEGMENTATION AFTER DEEP LEARNING PERFORM(blob all the sphere binary)
alldice=[]
acc=[]
Ts = [];
Data = [];

for ii=1:4
    subplot(1,4,ii)
    I = readimage(IMDS,ii);
    [C,scores] = semanticseg(I,net1);
    
    outt2=C=="foreground";
    st2=strel('disk',5);
    outt22=imerode(outt2,st2);
    %       outt22(:,:,ii) = imerode(outt2,st2);
    
    title('output')
    imshow(outt22)
    
    info = regionprops(outt22,'Boundingbox') ;
    imshow(outt22)
    hold on
    
    allbb=[]
    for k = 1 : length(info)
        BB = info(k).BoundingBox;
        rectangle('Position', [BB(1),BB(2),BB(3),BB(4)],'EdgeColor','r','LineWidth',2) ;
        allbb=[allbb; BB]
    end
    
    vol = [];
    for i=1:size(allbb,1)
        
        zz=imcrop(outt22,round(allbb(i,:)));
        
        stats = regionprops('table',zz,'Centroid',...
            'MajorAxisLength','MinorAxisLength');
        T = regionprops('table', outt22,'Area','Centroid')
        m1=stats.MajorAxisLength;
        m2=stats.MinorAxisLength;
        v1=stats.MinorAxisLength;
        
        vol(i)=m1*m2*(v1/2);
        
    end
    ss = 'the volume';
    for i = 1 : length(vol)
        ss=sprintf('%s %d is %6f ', ss, i, vol(i));
    end
    msgbox(ss);
end

[mohd akmal masud]

Its WORK sir..TQVM

Answer 2

[Voss]

You will have to adjust your code to handle cases when the number of regions in an image is not 4. In particular, the line

for i=1:4

can be changed to

for i = 1:length(info)

You will also need to modify your definition of ss for the same reason.