Hi, i need to divide all 5 images into 16x16 block, then perform "minus" operation (i.e 1st from 3rd image, then calculate "mean").Likewise for (3rd and 5th image) and (2nd and 4th image) using block process

Thanks in advance

Answer 1

[Walter Roberson]

See mat2cell() and cellfun()

Answer 2

[kaavya subramani]

Thanks a lot sir, i have tried code using those 2 function, please check whether it is correct.

n=16*ones(1,16)
k=[img1 img2 img3 img4 img5]
for i=1:length(k)
blockres(i)=mat2cell(i,n,n)
end
for p=1:length(blockres)-2
for q=2+p:length(blockres)
minus(p)=cellfun(@minus,p,q,'uniform',0)
end
sum(p)=cellfun(@sum,minus,'uniform',0)
end
mean(p)=cellfun(@mean,sum,'uniform',0)

[Walter Roberson]

In order for

k=[img1 img2 img3 img4 img5]

to work, each of img1, img2, and so on, would have to be a cell array. Remember, [] is horzcat(), so if img1, img2 and so on are images, that line of code would put all of them side by side as one larger image, and length(k) would probably end up as the number of columns in the combined image. You should be considering

k  = {img1, img2, img3, img4, img5};

Your line

blockres(i)=mat2cell(i,n,n)

would attempt to subdivide the scalar i into blocks of 16 by 16. You need to subdivide the i'th image instead. Such as k{i}

I did not check your cellfun logic.

[kaavya subramani]

oh k,k sir. Thanks,

so the code should be

n=16*ones(1,16)
k={img1 img2 img3 img4 img5}
for i=1:length(k)
blockres(i)=mat2cell(i,n,n)
end
for p=1:length(blockres)-2
for q=2+p:length(blockres)
minus(p)=cellfun(@minus,p,q,'uniform',0)
end
sum(p)=cellfun(@sum,minus,'uniform',0)
end
mean(p)=cellfun(@mean,sum,'uniform',0)

[Walter Roberson]

You are still trying to subdivide the scalar i instead of the i'th image. I already made a suggestion about how to fix that.

Your code

for q=2+p:length(blockres)
minus(p)=cellfun(@minus,p,q,'uniform',0)
end

tries to use p and q as cells with cellfun, but p and q are indexes, not cells. You need to use the indices to extract something from blockres, using {} notation.

That code also has the problem that it uses @minus to indicate that it wants to use the subtraction function, and then writes the result into a variable named minus. On the second iteration of the loop, minus would then refer to the variable and you would be attempting to take the handle of the variable, which is going to fail. You should ever make an assignment to a variable that is the same name as one of the important MATLAB functions.

[kaavya subramani]

Sorry sir,i noticed that from comment, but i not corrected in my code. n=16*ones(1,16) k={img1 img2 img3 img4 img5} for i=1:length(k) blockres(i)=mat2cell(k{i},n,n) end for p=1:length(blockres)-2 for q=2+p:length(blockres) minus(p)=cellfun(@minus,p{blockres},q{blockres},'uniform',0) end mean(p)=cellfun(@mean,sum,'uniform',0) end

[kaavya subramani]

k sir. n=16*ones(1,16) k={img1 img2 img3 img4 img5} for i=1:length(k) blockres(i)=mat2cell(k{i},n,n) end for p=1:length(blockres)-2 for q=2+p:length(blockres) res1(p)=cellfun(@minus,p{blockres},q{blockres},'uniform',0) end res2(p)=cellfun(@mean,sum,'uniform',0) end